About the Role

Anthropic is seeking a Certification Development Lead to build and scale a best-in-class certification program for our partner ecosystem. We’ve launched our first certification, the Claude Certified Architect, Foundations (CCA-F), and now we need someone to turn that first credential into a comprehensive, multi-tiered program that our systems integrator, consulting, and technology partners rely on to demonstrate real competence with Claude.

This is a 0-to-1 build. You’ll define the certification strategy, design the credential architecture, and build the program infrastructure from the ground up, all with our partner audience at the center. You’ll work at the intersection of program design and subject matter expertise: identifying what partners need to know, structuring the path to get them there, and collaborating with internal SMEs to ensure the content reflects how Claude actually works and where it’s headed. The systems and tooling that deliver certifications are being built by a separate team; your focus is the program itself, the credentialing framework, the partner experience, and the cross-functional relationships that keep it all connected.

The broader program strategy, including audience planning, partner relationships, and go-to-market alignment, lives within the Partner Enablement team. You’ll sit at the center of that work, collaborating closely with Partnerships, Applied AI, and GTM Productivity to ensure certifications serve both the partners earning them and the Anthropic teams relying on certified partners to extend our reach.

Responsibilities

• Own the end-to-end certification program: define the vision, build the credentialing framework, and establish the multi-tiered certification path that scales with Anthropic’s partner ecosystem
• Design certification architectures, competency frameworks, and assessment methodologies that measure real applied skill, not just content exposure
• Build with the partner audience in mind: understand the needs, workflows, and incentives of SI, consulting, and technology partners, and design credentialing experiences that meet them where they are
• Collaborate with internal subject matter experts across Applied AI, Product, and Engineering to translate Claude’s evolving capabilities into rigorous, current certification content
• Partner with trainers and to ensure alignment between training delivery and credentialing requirements, so the learning journey and credential path reinforce each other
• Work cross-functionally with Partnerships, GTM Productivity, and Revenue Operations to connect certification outcomes to partner performance, pipeline attribution, and ecosystem health metrics
• Define and manage relationships with credentialing vendors and collaborate with the systems owner on platform and delivery infrastructure decisions
• Establish quality standards for assessments including validity, reliability, and fairness, and use assessment data to continuously improve the program
• Build the operational processes around certification lifecycle management: development, launch, renewal, versioning, and retirement of credentials as Claude’s capabilities evolve

You May Be a Good Fit If You Have

• 7+ years in certification program design, credentialing, developer education, or partner enablement, with demonstrated experience building programs from scratch, not just maintaining existing ones
• A track record of 0-to-1 builds: you’ve stood up a certification or credentialing program where one didn’t exist before, defined the strategy, earned stakeholder buy-in, and shipped it
• Deep understanding of partner ecosystems, particularly SI and consulting partner audiences, and the ability to design certification experiences that drive real adoption and competency within those organizations
• Strong educational design instincts: you think in terms of competency frameworks, learning progressions, and what it actually means to demonstrate mastery versus having been exposed to content
• Experience with assessment methodologies beyond multiple-choice, including performance-based assessment, portfolio evaluation, or scenario-based testing that measures applied skill
• Baseline fluency with AI and technology products, with the ability and curiosity to learn Claude’s product suite deeply enough to articulate the “why” behind each certification tier and competency requirement
• Clear, structured communication and the ability to work with non-education stakeholders (partnerships, sales, engineering) while maintaining rigorous program standards
• Comfort operating in ambiguous, fast-moving environments, making decisions with incomplete information, and iterating quickly based on partner and internal feedback

Strong Candidates May Also Have

• Experience managing relationships with credentialing vendors (Pearson, Credly, or similar) and navigating the operational complexity of third-party certification delivery
• Familiarity with certification program economics: pricing, packaging, and the business model decisions that make a program sustainable
• Background in technical credentialing specifically, such as cloud certifications (AWS, GCP, Azure) or developer platform credentials
• Knowledge of psychometrics, credential validation, or assessment quality frameworks
• Experience benchmarking and learning from best-in-class programs from peer organizations
• Prior work at a high-growth technology company navigating rapid product change

Apply using this link. We are accepting applications on a rolling basis for the next cohort of Anthropic Fellows, which is expected to start in late September. In some circumstances, we can accommodate fellows starting outside the usual cohort timelines — please note in your application if the September start date doesn't work for you.

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Anthropic Fellows Program overview

The Anthropic Fellows Program is designed to foster AI research and engineering talent. We provide funding and mentorship to promising technical talent - regardless of previous experience.

Fellows will primarily use external infrastructure (e.g. open-source models, public APIs) to work on an empirical project aligned with our research priorities, with the goal of producing a public output (e.g. a paper submission). In one of our earlier cohorts, over 80% of fellows produced papers.

We run multiple cohorts of Fellows each year and review applications on a rolling basis. This application is for cohorts starting in July 2026 and beyond.

What to expect

• 4 months of full-time research 
• Direct mentorship from Anthropic researchers 
• Access to a shared workspace (in either Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (these vary by country)
• Funding for compute (~$15k/month) and other research expenses

Interview process

The interview process will include an initial application & reference check, technical assessments & interviews, and a research discussion. 

We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team.

Compensation

The expected base stipend for this role is 3,850 USD / 2,310 GBP / 4,300 CAD per week, with an expectation of 40 hours per week for 4 months (with possible extension).

Fellows workstreams

Due to the success of the Anthropic Fellows for AI Safety Research program, we are now expanding it across teams at Anthropic. We expect there to be significant overlap in the types of skills and responsibilities across the roles and will by default consider candidates for all the workstreams.

Some of the workstreams may include unique assessment steps; we therefore ask you for workstream preferences in the application. You can see an overview of the current workstreams below:

1. AI Safety Fellows
2. AI Security Fellows
3. ML Systems & Performance Fellows
4. Reinforcement Learning Fellows
5. Economics & Societal Impacts Fellows

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Across the workstreams, you may be a good fit if you:

• Are motivated by making sure AI is safe and beneficial for society as a whole
• Are excited to transition into empirical AI research and would be interested in a full-time role at Anthropic
• Have a strong technical background in computer science, mathematics, or physics
• Thrive in fast-paced, collaborative environments
• Can implement ideas quickly and communicate clearly

Strong candidates may also have:

• Strong background in a discipline relevant to a specific Fellows workstream (e.g. economics, social sciences, or cybersecurity)
• Experience in areas of research or engineering related to their workstream

Candidates must be:

• Fluent in Python programming
• Available to work full-time on the Fellows program

The Anthropic Institute Fellows (Economics & Policy)

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential research areas and mentors include:

• Economics
  • Maxim Massenkoff
  • Peter McCrory
• Policy, Security, and Society
  
  • Jack Clark
  • Marina Favaro
  • Jim Baker

Projects in this workstream may include:

• Designing and conducting empirical research on AI's economic effects, drawing on external data sources
• Developing new methodological approaches for studying AI's impact on labor markets, the future of work, and society
• Analysing the offense–defense balance for AI-enabled cyber and bio capabilities as models scale
• Measuring the extent to which model performance increases with custom harnesses?
• Identifying market driven mechanisms that could improve societal resilience to anticipated threats from AI systems?
• Identifying which metrics relating to AI R&D could service as early warning signals for recursive self-improvement

For past project examples, see:

• How AI Impacts Skill Formation: Judy Shen and Alex Tamkin
• Stress-Testing Model Specs Reveals Character Differences among Language Models: Jifan Zhang, Henry Sleight, Andi Peng, John Schulman, and Esin Durmus

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Have an interest in economics or policy research; prior experience in this area is a plus but not required
• Are adaptable and collaborative, able to take direction and contribute to team priorities rather than needing to pursue a predetermined research agenda
• Are skilled at writing up and communicating your results, even when they're null or unexpected
• Are passionate about translating research insights into actionable recommendations for improving AI systems and informing policy

Logistics

Logistics Requirements: To participate in the Fellows program, you must have work authorization in the US, UK, or Canada and be located in that country during the program.

Workspace Locations: We have designated shared workspaces in London and Berkeley where fellows will work from and mentors will visit. We are also open to remote fellows in the UK, US, or Canada. We will ask you about your availability to work from Berkeley or London (full- or part-time) during the program.

Visa Sponsorship: We are not currently able to sponsor visas for fellows. To participate in the Fellows program, you need to have or independently obtain full-time work authorization in the UK, the US, or Canada.

Program Duration: The program runs for 4 months, full-time. If you can't commit to the full duration, please still apply and note your constraints in the application. We review these requests on a case-by-case basis.

Please note: We do not guarantee that we will make any full-time offers to fellows. However, strong performance during the program may indicate that a Fellow would be a good fit for full-time roles at Anthropic. In previous cohorts, 25-50% of fellows received a full-time offer, and we’ve supported many more to go on to do great work on AI safety and security at other organizations.

Applications and interviews are managed by Constellation, our recruiting partner. Clicking "Apply here" will take you to their portal, and updates will come from a Constellation address. Constellation also runs the Berkeley workspace and provides program support for fellows working on AI safety and security; fellows on capabilities-focused projects are supported directly by Anthropic. All applicants currently use the same application portal but we are working to separate applications for safety/security and capabilities focused projects in future rounds.

Apply here

The below are Anthropic's policies for full time roles. These do NOT apply to the Fellows Program.

Apply using this link. We are accepting applications on a rolling basis for the next cohort of Anthropic Fellows, which is expected to start in late September. In some circumstances, we can accommodate fellows starting outside the usual cohort timelines — please note in your application if the September start date doesn't work for you.

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Anthropic Fellows Program overview

The Anthropic Fellows Program is designed to foster AI research and engineering talent. We provide funding and mentorship to promising technical talent - regardless of previous experience.

Fellows will primarily use external infrastructure (e.g. open-source models, public APIs) to work on an empirical project aligned with our research priorities, with the goal of producing a public output (e.g. a paper submission). In one of our earlier cohorts, over 80% of fellows produced papers.

We run multiple cohorts of Fellows each year and review applications on a rolling basis. This application is for cohorts starting in July 2026 and beyond.

What to expect

• 4 months of full-time research 
• Direct mentorship from Anthropic researchers 
• Access to a shared workspace (in either Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (these vary by country)
• Funding for compute (~$15k/month) and other research expenses

Interview process

The interview process will include an initial application & reference check, technical assessments & interviews, and a research discussion. 

We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team.

Compensation

The expected base stipend for this role is 3,850 USD / 2,310 GBP / 4,300 CAD per week, with an expectation of 40 hours per week for 4 months (with possible extension).

Fellows workstreams

Due to the success of the Anthropic Fellows for AI Safety Research program, we are now expanding it across teams at Anthropic. We expect there to be significant overlap in the types of skills and responsibilities across the roles and will by default consider candidates for all the workstreams.

Some of the workstreams may include unique assessment steps; we therefore ask you for workstream preferences in the application. You can see an overview of the current workstreams below:

1. AI Safety Fellows
2. AI Security Fellows
3. ML Systems & Performance Fellows
4. Reinforcement Learning Fellows
5. Economics & Societal Impacts Fellows

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Across the workstreams, you may be a good fit if you:

• Are motivated by making sure AI is safe and beneficial for society as a whole
• Are excited to transition into empirical AI research and would be interested in a full-time role at Anthropic
• Have a strong technical background in computer science, mathematics, or physics
• Thrive in fast-paced, collaborative environments
• Can implement ideas quickly and communicate clearly

Strong candidates may also have:

• Strong background in a discipline relevant to a specific Fellows workstream (e.g. economics, social sciences, or cybersecurity)
• Experience in areas of research or engineering related to their workstream

Candidates must be:

• Fluent in Python programming
• Available to work full-time on the Fellows program

Reinforcement Learning Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential research areas and mentors include:

• Ruhua Jiang
• Kaidi Cao
• Sunny Duan
• David Brandfonbrener
• Colt Steele
• Dino Distefano
• Will Williams

Projects in this workstream may include:

• Building model-based tools to better understand AI training data and improve training data quality
• A research project to better understand generalization
• Creating RL environments to improve Claude models at capabilities that are within your domain of expertise
• Building RL environments for safety-related tasks
• Conducting research and implementing solutions in areas such as RL algorithms

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Have strong software engineering skills with experience building complex ML systems
• Can balance research exploration with engineering rigor and operational reliability
• Enjoy collaborating across research and engineering disciplines
• Are comfortable working with large-scale distributed systems and high-performance computing
• Have experience with training, fine-tuning, or evaluating large language models
• Are adept at analyzing and debugging model training processes

Logistics

Logistics Requirements: To participate in the Fellows program, you must have work authorization in the US, UK, or Canada and be located in that country during the program.

Workspace Locations: We have designated shared workspaces in London and Berkeley where fellows will work from and mentors will visit. We are also open to remote fellows in the UK, US, or Canada. We will ask you about your availability to work from Berkeley or London (full- or part-time) during the program.

Visa Sponsorship: We are not currently able to sponsor visas for fellows. To participate in the Fellows program, you need to have or independently obtain full-time work authorization in the UK, the US, or Canada.

Program Duration: The program runs for 4 months, full-time. If you can't commit to the full duration, please still apply and note your constraints in the application. We review these requests on a case-by-case basis.

Please note: We do not guarantee that we will make any full-time offers to fellows. However, strong performance during the program may indicate that a Fellow would be a good fit for full-time roles at Anthropic. In previous cohorts, 25-50% of fellows received a full-time offer, and we’ve supported many more to go on to do great work on AI safety and security at other organizations.

Applications and interviews are managed by Constellation, our recruiting partner. Clicking "Apply here" will take you to their portal, and updates will come from a Constellation address. Constellation also runs the Berkeley workspace and provides program support for fellows working on AI safety and security; fellows on capabilities-focused projects are supported directly by Anthropic. All applicants currently use the same application portal but we are working to separate applications for safety/security and capabilities focused projects in future rounds.

Apply here

The below are Anthropic's policies for full time roles. These do NOT apply to the Fellows Program.

Apply using this link. We are accepting applications on a rolling basis for the next cohort of Anthropic Fellows, which is expected to start in late September. In some circumstances, we can accommodate fellows starting outside the usual cohort timelines — please note in your application if the September start date doesn't work for you.

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Anthropic Fellows Program overview

The Anthropic Fellows Program is designed to foster AI research and engineering talent. We provide funding and mentorship to promising technical talent - regardless of previous experience.

Fellows will primarily use external infrastructure (e.g. open-source models, public APIs) to work on an empirical project aligned with our research priorities, with the goal of producing a public output (e.g. a paper submission). In one of our earlier cohorts, over 80% of fellows produced papers.

We run multiple cohorts of Fellows each year and review applications on a rolling basis. This application is for cohorts starting in July 2026 and beyond.

What to expect

• 4 months of full-time research 
• Direct mentorship from Anthropic researchers 
• Access to a shared workspace (in either Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (these vary by country)
• Funding for compute (~$15k/month) and other research expenses

Interview process

The interview process will include an initial application & reference check, technical assessments & interviews, and a research discussion. 

We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team.

Compensation

The expected base stipend for this role is 3,850 USD / 2,310 GBP / 4,300 CAD per week, with an expectation of 40 hours per week for 4 months (with possible extension).

Fellows workstreams

Due to the success of the Anthropic Fellows for AI Safety Research program, we are now expanding it across teams at Anthropic. We expect there to be significant overlap in the types of skills and responsibilities across the roles and will by default consider candidates for all the workstreams.

Some of the workstreams may include unique assessment steps; we therefore ask you for workstream preferences in the application. You can see an overview of the current workstreams below:

1. AI Safety Fellows
2. AI Security Fellows
3. ML Systems & Performance Fellows
4. Reinforcement Learning Fellows
5. Economics & Societal Impacts Fellows

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Across the workstreams, you may be a good fit if you:

• Are motivated by making sure AI is safe and beneficial for society as a whole
• Are excited to transition into empirical AI research and would be interested in a full-time role at Anthropic
• Have a strong technical background in computer science, mathematics, or physics
• Thrive in fast-paced, collaborative environments
• Can implement ideas quickly and communicate clearly

Strong candidates may also have:

• Strong background in a discipline relevant to a specific Fellows workstream (e.g. economics, social sciences, or cybersecurity)
• Experience in areas of research or engineering related to their workstream

Candidates must be:

• Fluent in Python programming
• Available to work full-time on the Fellows program

ML Systems & Performance Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

• Alwin Peng
• Zygi Straznickas

Note: You may research mentors' prior work, but all applications must go through the official form, not the mentors.

For a past example of an engineering-heavy project, see:

• AI agents find $4.6M in blockchain smart contract exploits

Projects in this workstream may include:

• Building a CPU simulator for accelerator workloads
• Adding backends for different accelerators on an open source project
• Building on demand infrastructure for other infrastructure heavy fellows projects 
• Building complex synthetic data or environment pipelines

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Have strong software engineering skills with experience building complex ML systems
• Can balance research exploration with engineering rigor and operational reliability
• Enjoy collaborating across research and engineering disciplines
• Are comfortable working with large-scale distributed systems and high-performance computing (e.g. in trading)
• Have experience with training, fine-tuning, or evaluating large language models
• Are adept at analyzing and debugging model training processes

Logistics

Logistics Requirements: To participate in the Fellows program, you must have work authorization in the US, UK, or Canada and be located in that country during the program.

Workspace Locations: We have designated shared workspaces in London and Berkeley where fellows will work from and mentors will visit. We are also open to remote fellows in the UK, US, or Canada. We will ask you about your availability to work from Berkeley or London (full- or part-time) during the program.

Visa Sponsorship: We are not currently able to sponsor visas for fellows. To participate in the Fellows program, you need to have or independently obtain full-time work authorization in the UK, the US, or Canada.

Program Duration: The program runs for 4 months, full-time. If you can't commit to the full duration, please still apply and note your constraints in the application. We review these requests on a case-by-case basis.

Please note: We do not guarantee that we will make any full-time offers to fellows. However, strong performance during the program may indicate that a Fellow would be a good fit for full-time roles at Anthropic. In previous cohorts, 25-50% of fellows received a full-time offer, and we’ve supported many more to go on to do great work on AI safety and security at other organizations.

Applications and interviews are managed by Constellation, our recruiting partner. Clicking "Apply here" will take you to their portal, and updates will come from a Constellation address. Constellation also runs the Berkeley workspace and provides program support for fellows working on AI safety and security; fellows on capabilities-focused projects are supported directly by Anthropic. All applicants currently use the same application portal but we are working to separate applications for safety/security and capabilities focused projects in future rounds.

Apply here

The below are Anthropic's policies for full time roles. These do NOT apply to the Fellows Program.

Apply using this link. We are accepting applications on a rolling basis for the next cohort of Anthropic Fellows, which is expected to start in late September. In some circumstances, we can accommodate fellows starting outside the usual cohort timelines — please note in your application if the September start date doesn't work for you.

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Anthropic Fellows Program overview

The Anthropic Fellows Program is designed to foster AI research and engineering talent. We provide funding and mentorship to promising technical talent - regardless of previous experience.

Fellows will primarily use external infrastructure (e.g. open-source models, public APIs) to work on an empirical project aligned with our research priorities, with the goal of producing a public output (e.g. a paper submission). In one of our earlier cohorts, over 80% of fellows produced papers.

We run multiple cohorts of Fellows each year and review applications on a rolling basis. This application is for cohorts starting in July 2026 and beyond.

What to expect

• 4 months of full-time research 
• Direct mentorship from Anthropic researchers 
• Access to a shared workspace (in either Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (these vary by country)
• Funding for compute (~$15k/month) and other research expenses

Interview process

The interview process will include an initial application & reference check, technical assessments & interviews, and a research discussion. 

We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team.

Compensation

The expected base stipend for this role is 3,850 USD / 2,310 GBP / 4,300 CAD per week, with an expectation of 40 hours per week for 4 months (with possible extension).

Fellows workstreams

Due to the success of the Anthropic Fellows for AI Safety Research program, we are now expanding it across teams at Anthropic. We expect there to be significant overlap in the types of skills and responsibilities across the roles and will by default consider candidates for all the workstreams.

Some of the workstreams may include unique assessment steps; we therefore ask you for workstream preferences in the application. You can see an overview of the current workstreams below:

1. AI Safety Fellows
2. AI Security Fellows
3. ML Systems & Performance Fellows
4. Reinforcement Learning Fellows
5. Economics & Societal Impacts Fellows

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Across the workstreams, you may be a good fit if you:

• Are motivated by making sure AI is safe and beneficial for society as a whole
• Are excited to transition into empirical AI research and would be interested in a full-time role at Anthropic
• Have a strong technical background in computer science, mathematics, or physics
• Thrive in fast-paced, collaborative environments
• Can implement ideas quickly and communicate clearly

Strong candidates may also have:

• Strong background in a discipline relevant to a specific Fellows workstream (e.g. economics, social sciences, or cybersecurity)
• Experience in areas of research or engineering related to their workstream

Candidates must be:

• Fluent in Python programming
• Available to work full-time on the Fellows program

AI Security Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

• Nicholas Carlini
• Keri Warr
• Evyatar Ben Asher
• Keane Lucas
• Newton Cheng

On our Alignment Science and Frontier Red Team blogs, you can read about some past Fellows projects, including:

• AI agents find $4.6M in blockchain smart contract exploits: Winnie Xiao and Cole Killian, mentored by Nicholas Carlini and Alwin Peng
• Strengthening Red Teams: A Modular Scaffold for Control Evaluations: Chloe Loughridge et al., mentored by Jon Kutasov and Joe Benton

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Are motivated by reducing catastrophic risks from advanced AI systems
• Have contributed to open-source projects in LLM- or security-adjacent repositories
• Have demonstrated success in bringing clarity and ownership to ambiguous technical problems
• Have experience with pentesting, vulnerability research, or other offensive security work
• Have a demonstrated willingness to do the "dirty work" that produces high-quality outputs
• Have reported CVEs or been awarded bug bounties
• Have experience with empirical ML research projects
• Have experience with deep learning frameworks and experiment management

Logistics

Logistics Requirements: To participate in the Fellows program, you must have work authorization in the US, UK, or Canada and be located in that country during the program.

Workspace Locations: We have designated shared workspaces in London and Berkeley where fellows will work from and mentors will visit. We are also open to remote fellows in the UK, US, or Canada. We will ask you about your availability to work from Berkeley or London (full- or part-time) during the program.

Visa Sponsorship: We are not currently able to sponsor visas for fellows. To participate in the Fellows program, you need to have or independently obtain full-time work authorization in the UK, the US, or Canada.

Program Duration: The program runs for 4 months, full-time. If you can't commit to the full duration, please still apply and note your constraints in the application. We review these requests on a case-by-case basis.

Please note: We do not guarantee that we will make any full-time offers to fellows. However, strong performance during the program may indicate that a Fellow would be a good fit for full-time roles at Anthropic. In previous cohorts, 25-50% of fellows received a full-time offer, and we’ve supported many more to go on to do great work on AI safety and security at other organizations.

Applications and interviews are managed by Constellation, our recruiting partner. Clicking "Apply here" will take you to their portal, and updates will come from a Constellation address. Constellation also runs the Berkeley workspace and provides program support for fellows working on AI safety and security; fellows on capabilities-focused projects are supported directly by Anthropic. All applicants currently use the same application portal but we are working to separate applications for safety/security and capabilities focused projects in future rounds.

Apply here

The below are Anthropic's policies for full time roles. These do NOT apply to the Fellows Program.

Apply using this link. We are accepting applications on a rolling basis for the next cohort of Anthropic Fellows, which is expected to start in late September. In some circumstances, we can accommodate fellows starting outside the usual cohort timelines — please note in your application if the September start date doesn't work for you.

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Anthropic Fellows Program overview

The Anthropic Fellows Program is designed to foster AI research and engineering talent. We provide funding and mentorship to promising technical talent - regardless of previous experience.

Fellows will primarily use external infrastructure (e.g. open-source models, public APIs) to work on an empirical project aligned with our research priorities, with the goal of producing a public output (e.g. a paper submission). In one of our earlier cohorts, over 80% of fellows produced papers.

We run multiple cohorts of Fellows each year and review applications on a rolling basis. This application is for cohorts starting in July 2026 and beyond.

What to expect

• 4 months of full-time research 
• Direct mentorship from Anthropic researchers 
• Access to a shared workspace (in either Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (these vary by country)
• Funding for compute (~$15k/month) and other research expenses

Interview process

The interview process will include an initial application & reference check, technical assessments & interviews, and a research discussion. 

We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team.

Compensation

The expected base stipend for this role is 3,850 USD / 2,310 GBP / 4,300 CAD per week, with an expectation of 40 hours per week for 4 months (with possible extension).

Fellows workstreams

Due to the success of the Anthropic Fellows for AI Safety Research program, we are now expanding it across teams at Anthropic. We expect there to be significant overlap in the types of skills and responsibilities across the roles and will by default consider candidates for all the workstreams.

Some of the workstreams may include unique assessment steps; we therefore ask you for workstream preferences in the application. You can see an overview of the current workstreams below:

1. AI Safety Fellows
2. AI Security Fellows
3. ML Systems & Performance Fellows
4. Reinforcement Learning Fellows
5. Economics & Societal Impacts Fellows

This page is specific to one of the Anthropic Fellows Workstreams, see also the main Anthropic Fellows posting.

Across the workstreams, you may be a good fit if you:

• Are motivated by making sure AI is safe and beneficial for society as a whole
• Are excited to transition into empirical AI research and would be interested in a full-time role at Anthropic
• Have a strong technical background in computer science, mathematics, or physics
• Thrive in fast-paced, collaborative environments
• Can implement ideas quickly and communicate clearly

Strong candidates may also have:

• Strong background in a discipline relevant to a specific Fellows workstream (e.g. economics, social sciences, or cybersecurity)
• Experience in areas of research or engineering related to their workstream

Candidates must be:

• Fluent in Python programming
• Available to work full-time on the Fellows program

AI Safety Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

• Sam Bowman
• Sara Price
• Alex Tamkin
• Nina Panickssery
• Trenton Bricken
• Logan Graham
• Jascha Sohl-Dickstein
• Joe Benton
• Collin Burns
• Fabien Roger
• Samuel Marks
• Kyle Fish
• Ethan Perez

Our mentors will lead projects in select AI safety research areas, such as:

• Scalable Oversight: Developing techniques to keep highly capable models helpful and honest, even as they surpass human-level intelligence in various domains.
• Adversarial Robustness and AI Control: Creating methods to ensure advanced AI systems remain safe and harmless in unfamiliar or adversarial scenarios.
• Model Organisms: Creating model organisms of misalignment to improve our empirical understanding of how alignment failures might arise.
• Model Internals / Mechanistic Interpretability: Advancing our understanding of the internal workings of large language models to enable more targeted interventions and safety measures.
• AI Welfare: Improving our understanding of potential AI welfare and developing related evaluations and mitigations.

On our Alignment Science and Frontier Red Team blogs, you can read about past projects, including:

• Subliminal Learning: Language Models Transmit Behavioral Traits via Hidden Signals in Data: Alex Cloud and Minh Le, et al., mentors including Samuel Marks and Owain Evans
• Open-source circuits: Michael Hanna and Mateusz Piotrowski with mentorship from Emmanuel Ameisen and Jack Lindsey

For a full list of representative projects for each area, please see these blog posts: Introducing the Anthropic Fellows Program for AI Safety Research, Recommendations for Technical AI Safety Research Directions.

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Are motivated by reducing catastrophic risks from advanced AI systems
• Have experience with empirical ML research projects
• Have experience working with large language models
• Have experience in one of the research areas mentioned above
• Have a track record of open-source contributions

Logistics

Logistics Requirements: To participate in the Fellows program, you must have work authorization in the US, UK, or Canada and be located in that country during the program.

Workspace Locations: We have designated shared workspaces in London and Berkeley where fellows will work from and mentors will visit. We are also open to remote fellows in the UK, US, or Canada. We will ask you about your availability to work from Berkeley or London (full- or part-time) during the program.

Visa Sponsorship: We are not currently able to sponsor visas for fellows. To participate in the Fellows program, you need to have or independently obtain full-time work authorization in the UK, the US, or Canada.

Program Duration: The program runs for 4 months, full-time. If you can't commit to the full duration, please still apply and note your constraints in the application. We review these requests on a case-by-case basis.

Please note: We do not guarantee that we will make any full-time offers to fellows. However, strong performance during the program may indicate that a Fellow would be a good fit for full-time roles at Anthropic. In previous cohorts, 25-50% of fellows received a full-time offer, and we’ve supported many more to go on to do great work on AI safety and security at other organizations.

Applications and interviews are managed by Constellation, our recruiting partner. Clicking "Apply here" will take you to their portal, and updates will come from a Constellation address. Constellation also runs the Berkeley workspace and provides program support for fellows working on AI safety and security; fellows on capabilities-focused projects are supported directly by Anthropic. All applicants currently use the same application portal but we are working to separate applications for safety/security and capabilities focused projects in future rounds.

Apply here

The below are Anthropic's policies for full time roles. These do NOT apply to the Fellows Program.

Apply using this link. We are accepting applications on a rolling basis for the next cohort of Anthropic Fellows, which is expected to start in late September. In some circumstances, we can accommodate fellows starting outside the usual cohort timelines — please note in your application if the September start date doesn't work for you.

Anthropic Fellows Program overview

The Anthropic Fellows Program is designed to foster AI research and engineering talent. We provide funding and mentorship to promising technical talent - regardless of previous experience.

Fellows will primarily use external infrastructure (e.g. open-source models, public APIs) to work on an empirical project aligned with our research priorities, with the goal of producing a public output (e.g. a paper submission). In one of our earlier cohorts, over 80% of fellows produced papers.

We run multiple cohorts of Fellows each year and review applications on a rolling basis. This application is for cohorts starting in July 2026 and beyond.

What to expect

• 4 months of full-time research 
• Direct mentorship from Anthropic researchers 
• Access to a shared workspace (in either Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (these vary by country)
• Funding for compute (~$15k/month) and other research expenses

Interview process

The interview process will include an initial application & reference check, technical assessments & interviews, and a research discussion. 

We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team.

Compensation

The expected base stipend for this role is 3,850 USD / 2,310 GBP / 4,300 CAD per week, with an expectation of 40 hours per week for 4 months (with possible extension).

Fellows workstreams

Due to the success of the Anthropic Fellows for AI Safety Research program, we are now expanding it across teams at Anthropic. We expect there to be significant overlap in the types of skills and responsibilities across the roles and will by default consider candidates for all the workstreams.

Some of the workstreams may include unique assessment steps; we therefore ask you for workstream preferences in the application. You can see an overview of the current workstreams below:

1. AI Safety Fellows
2. AI Security Fellows
3. ML Systems & Performance Fellows
4. Reinforcement Learning Fellows
5. Economics Fellows

Across the workstreams, you may be a good fit if you:

• Are motivated by making sure AI is safe and beneficial for society as a whole
• Are excited to transition into empirical AI research and would be interested in a full-time role at Anthropic
• Have a strong technical background in computer science, mathematics, or physics
• Thrive in fast-paced, collaborative environments
• Can implement ideas quickly and communicate clearly

Strong candidates may also have:

• Strong background in a discipline relevant to a specific Fellows workstream (e.g. economics, social sciences, or cybersecurity)
• Experience in areas of research or engineering related to their workstream

Candidates must be:

• Fluent in Python programming
• Available to work full-time on the Fellows program

AI Safety Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

• Sam Bowman
• Sara Price
• Alex Tamkin
• Nina Panickssery
• Trenton Bricken
• Logan Graham
• Jascha Sohl-Dickstein
• Joe Benton
• Fabien Roger
• Samuel Marks
• Kyle Fish
• Ethan Perez

Note: You may research mentors' prior work, but all applications must go through the official form, not the mentors.

Our mentors will lead projects in select AI safety research areas, such as:

• Scalable Oversight: Developing techniques to keep highly capable models helpful and honest, even as they surpass human-level intelligence in various domains.
• Adversarial Robustness and AI Control: Creating methods to ensure advanced AI systems remain safe and harmless in unfamiliar or adversarial scenarios.
• Model Organisms: Creating model organisms of misalignment to improve our empirical understanding of how alignment failures might arise.
• Model Internals / Mechanistic Interpretability: Advancing our understanding of the internal workings of large language models to enable more targeted interventions and safety measures.
• AI Welfare: Improving our understanding of potential AI welfare and developing related evaluations and mitigations.

On our Alignment Science and Frontier Red Team blogs, you can read about past projects, including:

• Subliminal Learning: Language Models Transmit Behavioral Traits via Hidden Signals in Data: Alex Cloud and Minh Le, et al., mentors including Samuel Marks and Owain Evans
• Open-source circuits: Michael Hanna and Mateusz Piotrowski with mentorship from Emmanuel Ameisen and Jack Lindsey

For a full list of representative projects for each area, please see these blog posts: Introducing the Anthropic Fellows Program for AI Safety Research, Recommendations for Technical AI Safety Research Directions.

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Are motivated by reducing catastrophic risks from advanced AI systems
• Have experience with empirical ML research projects
• Have experience working with large language models
• Have experience in one of the research areas mentioned above
• Have a track record of open-source contributions

AI Security Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

• Nicholas Carlini
• Keri Warr
• Evyatar Ben Asher
• Keane Lucas
• Newton Cheng

On our Alignment Science and Frontier Red Team blogs, you can read about some past Fellows projects, including:

• AI agents find $4.6M in blockchain smart contract exploits: Winnie Xiao and Cole Killian, mentored by Nicholas Carlini and Alwin Peng
• Strengthening Red Teams: A Modular Scaffold for Control Evaluations: Chloe Loughridge et al., mentored by Jon Kutasov and Joe Benton

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Are motivated by reducing catastrophic risks from advanced AI systems
• Have contributed to open-source projects in LLM- or security-adjacent repositories
• Have demonstrated success in bringing clarity and ownership to ambiguous technical problems
• Have experience with pentesting, vulnerability research, or other offensive security work
• Have a demonstrated willingness to do the "dirty work" that produces high-quality outputs
• Have reported CVEs or been awarded bug bounties
• Have experience with empirical ML research projects
• Have experience with deep learning frameworks and experiment management

ML Systems & Performance Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

• Alwin Peng
• Zygi Straznickas

For a past example of an engineering-heavy project, see:

• AI agents find $4.6M in blockchain smart contract exploits

Projects in this workstream may include:

• Building a CPU simulator for accelerator workloads
• Adding backends for different accelerators on an open source project
• Building on demand infrastructure for other infrastructure heavy fellows projects 
• Building complex synthetic data or environment pipelines

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Have strong software engineering skills with experience building complex ML systems
• Can balance research exploration with engineering rigor and operational reliability
• Enjoy collaborating across research and engineering disciplines
• Are comfortable working with large-scale distributed systems and high-performance computing (e.g. in trading)
• Have experience with training, fine-tuning, or evaluating large language models
• Are adept at analyzing and debugging model training processes

Reinforcement Learning Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential research areas and mentors include:

• Ruhua Jiang
• Kaidi Cao
• Sunny Duan
• David Brandfonbrener
• Colt Steele
• Dino Distefano
• Will Williams

Projects in this workstream may include:

• Building model-based tools to better understand AI training data and improve training data quality
• A research project to better understand generalization
• Creating RL environments to improve Claude models at capabilities that are within your domain of expertise
• Building RL environments for safety-related tasks
• Conducting research and implementing solutions in areas such as RL algorithms

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Have strong software engineering skills with experience building complex ML systems
• Can balance research exploration with engineering rigor and operational reliability
• Enjoy collaborating across research and engineering disciplines
• Are comfortable working with large-scale distributed systems and high-performance computing
• Have experience with training, fine-tuning, or evaluating large language models
• Are adept at analyzing and debugging model training processes

The Anthropic Institute Fellows (Economics & Policy)

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential research areas and mentors include:

• Economics
  • Maxim Massenkoff
  • Peter McCrory
• Policy, Security, and Society
  
  • Jack Clark
  • Marina Favaro
  • Jim Baker

Projects in this workstream may include:

• Designing and conducting empirical research on AI's economic effects, drawing on external data sources
• Developing new methodological approaches for studying AI's impact on labor markets, the future of work, and society
• Analysing the offense–defense balance for AI-enabled cyber and bio capabilities as models scale
• Measuring the extent to which model performance increases with custom harnesses?
• Identifying market driven mechanisms that could improve societal resilience to anticipated threats from AI systems?
• Identifying which metrics relating to AI R&D could service as early warning signals for recursive self-improvement

For past project examples, see:

• How AI Impacts Skill Formation: Judy Shen and Alex Tamkin
• Stress-Testing Model Specs Reveals Character Differences among Language Models: Jifan Zhang, Henry Sleight, Andi Peng, John Schulman, and Esin Durmus

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

• Have an interest in economics or policy research; prior experience in this area is a plus but not required
• Are adaptable and collaborative, able to take direction and contribute to team priorities rather than needing to pursue a predetermined research agenda
• Are skilled at writing up and communicating your results, even when they're null or unexpected
• Are passionate about translating research insights into actionable recommendations for improving AI systems and informing policy

Logistics

Logistics Requirements: To participate in the Fellows program, you must have work authorization in the US, UK, or Canada and be located in that country during the program.

Workspace Locations: We have designated shared workspaces in London and Berkeley where fellows will work from and mentors will visit. We are also open to remote fellows in the UK, US, or Canada. We will ask you about your availability to work from Berkeley or London (full- or part-time) during the program.

Visa Sponsorship: We are not currently able to sponsor visas for fellows. To participate in the Fellows program, you need to have or independently obtain full-time work authorization in the UK, the US, or Canada.

Program Duration: The program runs for 4 months, full-time. If you can't commit to the full duration, please still apply and note your constraints in the application. We review these requests on a case-by-case basis.

Please note: We do not guarantee that we will make any full-time offers to fellows. However, strong performance during the program may indicate that a Fellow would be a good fit for full-time roles at Anthropic. In previous cohorts, 25-50% of fellows received a full-time offer, and we’ve supported many more to go on to do great work on AI safety and security at other organizations.

Applications and interviews are managed by Constellation, our recruiting partner. Clicking "Apply here" will take you to their portal, and updates will come from a Constellation address. Constellation also runs the Berkeley workspace and provides program support for fellows working on AI safety and security; fellows on capabilities-focused projects are supported directly by Anthropic. All applicants currently use the same application portal but we are working to separate applications for safety/security and capabilities focused projects in future rounds.

Apply here

The below are Anthropic's policies for full time roles. These do NOT apply to the Fellows Program.

About the role

The Knowledge Work team builds the training environments and evaluations that make Claude effective at real-world professional workflows — searching, analyzing, and creating across the tools and documents knowledge workers use every day. As that work scales, the systems behind it need to be as rigorous as the research itself.

We are looking for a Research Engineer to own the reliability, observability, and infrastructure foundation that the team's research depends on. You will be responsible for ensuring our training and evaluation runs remain stable, well-instrumented, and high-quality as they grow in scale and complexity.

A core part of this role is shifting reliability work from reactive to proactive: hardening systems, stress-testing at realistic scale, and building the observability and tooling that surface problems early — so researchers can stay focused on research rather than incident response. You will be the team's stable, context-rich owner for environment health and evaluation integrity, and the primary point of contact for partner teams when issues arise.

Where this role focuses: While you'll work closely with researchers building new training environments, the priority for this role is the reliability those environments depend on. It's best suited to an engineer who finds real ownership and impact in making critical systems dependable, and in being the person behind trustworthy evaluation results the entire organization relies on.

Key Responsibilities:

• Serve as the dedicated reliability owner for the Knowledge Work training environments, providing continuity of context and reducing the operational overhead of rotating ownership

• Own a clean, canonical set of evaluation tools and processes for Knowledge Work capabilities, including the process used for model releases

• Build and automate observability, dashboards, and operational tooling for our training environments and evaluation systems, with an emphasis on high signal-to-noise: a small set of trusted metrics and alerts rather than sprawling instrumentation

• Proactively harden environments and evaluation systems through load testing, fault injection, and stress testing at realistic scale, so failures surface early rather than during critical training work

• Act as the primary point of contact for partner training and infrastructure teams when issues in our environments arise, and drive incidents to resolution

• Reduce the operational burden on researchers so they can stay focused on research

Minimum Qualifications:

• Highly experienced Python engineer who ships reliable, well-instrumented code that teammates trust in production

• Demonstrated experience operating ML or distributed systems at scale, including significant on-call and incident-response experience

• Strong SRE or production-engineering mindset — reaching for SLOs, load tests, and failure injection before reaching for more dashboards

• Foundational ML knowledge sufficient to understand what a training environment or evaluation is actually measuring, and recognize when an evaluation has become stale or gameable

• Able to read research code and reason evaluation integrity

Preferred Qualifications:

• 5+ years of experience operating ML or distributed systems at scale

• Experience building or operating RL environments, agent harnesses, or LLM evaluation frameworks

• Familiarity with reward modeling, evaluation design, or detecting and mitigating reward hacking

• Experience with bservability stacks (metrics, tracing, structured logging) and operational dashboard tooling

• Background in chaos engineering, fault injection, or large-scale load testing

• Experience with data quality pipelines, drift detection, or evaluation-set curation and versioning

• Familiarity with large-scale training or inference infrastructure (schedulers, multi-agent orchestration, sandboxed execution)

• Prior experience as a dedicated reliability or operations owner embedded within a research team

Responsibilities:

• Designing and implementing from scratch new information architecture strategies
• Performing finetuning and reinforcement learning to teach language models how to interact with new information architectures
• Building “hard” knowledge base eval sets to help identify failure modes of how language models work with external data
• Designing and evaluating advanced agentic search capabilities.

You may be a good fit if you:

• Are a very experienced Python programmer who can quickly produce reliable, high quality code that your teammates love using
• Have good machine learning research experience 
• Have experience developing software that utilizes Large Language Models such as Claude
• Are results-oriented, with a bias towards flexibility and impact
• Pick up slack, even if it goes outside your job description
• Enjoy pair programming (we love to pair!)
• Want to partner with world-class ML researchers to develop new LLM capabilities
• Care about the societal impacts of your work
• Have clear written and verbal communication

Strong candidates will also have experience with:

• Collaborating with product teams to quickly prototype and deliver innovative solutions
• Building complex agentic systems that utilize LLMs
• Developing scalable distributed information retrieval systems, such as search engines, knowledge graphs, RAG, indexing, ranking, query understanding, and distributed data processing

About the role

Anthropic's Human Data team partners with external domain experts — clinical pharmacists, securities lawyers, physicists, philosophers, and more — whose work directly shapes what Claude can do and how safely it does it.

We're looking for a Program Operations Manager to build and run the operational backbone of this program. You'll own the systems that keep a large, distributed group of external experts productive: onboarding, access and tooling, communications, data integrity across our vendor partner, and the reporting layer that shows how the program is performing.

The processes you'll run are still being built, and you'll be executing them at a volume that surfaces gaps nobody has seen yet. A meaningful part of the role is noticing what's going poorly, improving or automating it, and feeding that back into how the program is designed. You'll be the person research teams come to when they need experts ramped, and the person experts come to when they're blocked. It's a highly visible role with a clear throughline to the work that trains Claude.

Responsibilities: 

• Run onboarding end-to-end. Own the workflow that takes external experts from offer-accepted to productive — setup, access, communications, and the things that inevitably go sideways.

• Own access provisioning. Grant, audit, and maintain access across collaboration tools and internal systems for a large external population. Build the documented, auditable process that makes this scale.

• Maintain program data integrity. Keep the tracking systems research teams rely on accurate and current; reconcile records with our external workforce partner and catch discrepancies early.

• Build program reporting. Stand up weekly metrics on headcount, time-to-productive, support volume, and domain coverage. Use the data to spot gaps and surface recommendations.

• Partner across the organization. Work with research DRIs, IT, and our vendor partner to resolve access issues and keep experts unblocked.

You may be a good fit if you: 

• Have 4+ years of experience in program operations, research operations, recruiting coordination, vendor management, or a similar operational role

• Have owned an operational process end-to-end — you've been the person accountable when something broke, and the person who fixed it

• Are comfortable being the primary point of contact for a large, distributed population — you answer questions directly rather than routing them

• Exercise strong judgment about prioritization when the volume of work exceeds the time available, and know when to escalate versus resolve independently

• Are highly detail-oriented and take real ownership of data accuracy — you notice when a tracker is stale and can't leave it that way

• Write clearly and concisely; most of this job happens over Slack and email, and clarity compounds across hundreds of interactions

• Are comfortable executing at high volume in a program that's still being shaped — you handle edge cases well and surface what you learn

• Use AI tools as a core part of how you work — drafting communications, cleaning data, building lightweight automations — and are comfortable with spreadsheets when that's the right tool

Strong candidates may also have: 

• Experience scaling an operational program through rapid growth

• Background in external workforce, vendor, or recruiting operations

• Light scripting or automation experience (Apps Script, Zapier, or similar)

About the Role

As a Research Engineer / Scientist on the Societal Impacts team, you'll design and build critical infrastructure that enables and accelerates foundational research into how our AI systems impact people and society. Your work will directly contribute to our research publications, policy campaigns, safety systems, and products. 

Our team combines rigorous empirical methods with creative technical approaches. We’re currently grappling with big questions on how AI might impact the future of work, people's wellbeing, education, and more. Additionally, we are continuously studying socio-technical alignment (what values do our systems have?), and evaluating novel AI capabilities as they arise. We develop privacy-preserving tools to measure AI's effects at scale, conduct mixed-methods studies of human-AI interaction, and translate research insights into actionable recommendations for both product and policy. You can learn more about our team here

Strong candidates will have a track record of running & designing experiments relating to machine learning systems, building data processing pipelines, architecting & implementing high-quality internal infrastructure, working in a fast-paced startup environment, navigating the ambiguity inherent to novel empirical research, and demonstrating an eagerness to develop their own research & technical skills. The ideal candidate will enjoy a mixture of running experiments, developing new tools & evaluation suites, working cross-functionally across multiple research and product teams, and striving for beneficial & safe uses for AI.

Responsibilities:

• Design and implement scalable technical infrastructure that enables researchers to efficiently run experiments and evaluate AI systems.
• Architect systems that can handle uncertain and changing requirements while maintaining high standards of reliability
• Lead technical design discussions to ensure our infrastructure can support both current needs and future research directions
• Partner closely with researchers, data scientists, policy experts, and other cross-functional partners to advance Anthropic’s safety mission
• Interface with and improve our internal technical infrastructure and tools
• Generate net-new insights about the potential societal impact of systems being developed by Anthropic
• Ship changes that help improve our models and products based on the empirical research the Societal Impacts team is conducting 

Minimum qualifications

• Experience working ambiguous AI research projects
• Experience building and maintaining production-grade internal tools or research infrastructure
• Take pride in writing clean, well-documented code in Python that others can build upon
• Comfortable making technical decisions with incomplete information while maintaining high engineering standards
• Comfortable getting up-to-speed quickly on unfamiliar codebases, and can work well with other engineers with different backgrounds across the organization
• Track record of using technical infrastructure to interface effectively with machine learning models
• Have experience deriving insights from imperfect data streams
• Experience building systems and products on top of LLMs

Preferred qualifications

• Experience maintaining large, foundational infrastructure
• Experience incubating and maturing tooling platforms used by a wide variety of stakeholders
• Experience building simple interfaces that allow non-technical collaborators to evaluate AI systems
• Experience working with and prioritizing requests from a wide variety of stakeholders, including research and product teams
• Experience with distributed systems and designing for scale and reliability
• Experience scaling and optimizing the performance of tools
• Experience with product management and/or full-stack product engineering, with a track record of zero-to-one work in startup (or startup-like environments) and owning products end-to-end
• Research relating to the societal impacts of technology

Representative Projects:

• Design, implement, and mature scalable infrastructure for running large-scale experiments on how people interact with our AI systems (like Clio, Anthropic Interviewer, the Economic Index, Values in the Wild, and Emotional Impacts). 
• Build robust monitoring systems that help us detect and understand potential misuse or unexpected behaviors
• Create internal tools that help researchers, policy experts, and product teams quickly analyze dynamically changing AI system characteristics

About the role

Every request that hits Claude — from claude.ai, the API, our cloud partners, or internal research — passes through a routing decision. Not a generic load balancer round-robin, but a decision that accounts for what's already cached where, which accelerator the request runs best on, and what else is in flight across the fleet. Get it right and you extract meaningfully more throughput from the same hardware. Get it wrong and you burn capacity, miss latency SLOs, or shed load that shouldn't have been shed.

The Inference Routing team owns this layer. We build the cluster-level routing and coordination plane for Anthropic's inference fleet — the system that sits between the API surface and the inference engines themselves, making fleet-wide efficiency decisions in real time. As Anthropic moves from "many independent inference replicas" toward "a single warehouse-scale computer running a coordinated program," Dystro is the coordination layer.

This is a deeply technical team. The engineers here design custom load-balancing algorithms, build quantitative models of system performance, debug latency spikes that cross kernel, network, and framework boundaries, and reason carefully about cache placement across thousands of accelerators. They work shoulder-to-shoulder with teams that write kernels and ML framework internals. The EM for this team doesn't need to write kernels — but they do need the systems depth to make architectural calls, evaluate deeply technical candidates, and spot when a proposed optimization will have second-order effects on the fleet.

You'll inherit a strong team of distributed-systems engineers, and you'll be accountable for two things that pull in different directions: shipping system-level performance improvements that measurably increase fleet throughput and efficiency, and running the team operationally so that deploys are safe, incidents are rare, and the teams who depend on Dystro can plan around you with confidence. The job is holding both.

Representative work:

Things the Inference Routing EM actually spends time on:

• Deciding whether a proposed routing algorithm change is worth the deploy risk, given the modeled throughput gain and the blast radius if it regresses
• Sequencing a quarter where KV-cache offload, a new coordination protocol, and two model launches all compete for the same engineers
• Working through a persistent tail-latency regression with the team — walking down from fleet-level metrics to per-replica behavior to a root cause in the networking stack
• Building the case (with numbers) to peer teams for why a cross-team protocol change unlocks the next efficiency win
• Running the post-incident review after a cache-eviction bug caused a capacity event, and turning it into process changes that stick
• Interviewing a candidate who has built schedulers at supercomputing scale, and deciding whether they'd be additive to a team that already goes deep

What you'll do:

Drive system-level performance

• Own the technical roadmap for cluster-level inference efficiency — routing decisions, cache placement and eviction, cross-replica coordination, and the protocols that keep routing and inference engines in sync
• Partner with the inference engine, kernels, and performance teams to identify fleet-level throughput and latency wins, then turn those into shipped improvements with measurable results
• Build the team's habit of quantitative performance modeling: claim a win only when you can measure it, and know before you ship what the expected effect is

Deliver reliably and operate cleanly

• Set technical strategy for how routing evolves across heterogeneous hardware (GPUs, TPUs, Trainium) and across all our serving surfaces
• Run the team's operational backbone — on-call rotation, incident response, postmortem review, deploy safety — so the team can ship aggressively without the system becoming fragile
• Create clarity at a seam: Inference Routing sits between the API surface, the inference engines, and the cloud deployment teams. You'll make sure commitments are realistic, dependencies are understood, and nobody is surprised

Build and grow the team

• Develop and retain a strong existing team, and hire against the bar described above: people who can go to the OS and framework level when the problem demands it, and who care about production reliability
• Coach engineers through a roadmap where priorities shift with model launches, new hardware, and scaling demands. We pair a lot here — you'll help make that collaboration pattern productive
• Pick up slack when it matters. This is a small team in a critical path; sometimes the EM is the one unblocking a stuck deploy or synthesizing a design debate

You may be a good fit if you:

• Have 5+ years of engineering management experience, ideally with at least part of that leading teams on critical-path production infrastructure at scale
• Have a deep systems background — load balancing, scheduling, cache-coherent distributed state, high-performance networking, or similar. You need enough depth to make architectural calls about routing and efficiency, and to evaluate candidates who go to the kernel and framework level
• Have shipped performance improvements in large-scale systems and can explain, with numbers, what the impact was
• Have run production infrastructure with real operational stakes: on-call, incident response, capacity events, deploy discipline
• Are results-oriented with a bias toward impact, and comfortable working in a space where throughput, latency, stability, and feature velocity all pull in different directions
• Build strong relationships across team boundaries — this is a seam role, and much of the job is making sure other teams can rely on yours
• Are curious about machine learning systems. You don't need an ML research background, but you should want to learn how transformer inference actually works and how that shapes the systems problems

Strong candidates may also have:

• Experience with LLM inference serving — KV caching, continuous batching, request scheduling, prefill/decode disaggregation
• Background in cluster schedulers, load balancers, service meshes, or coordination planes at scale
• Familiarity with heterogeneous accelerator fleets (GPU/TPU/Trainium) and how hardware differences affect workload placement
• Experience with GPU/accelerator programming, ML framework internals, or OS-level performance debugging — enough to follow and evaluate the technical work, not necessarily to do it daily
• Led teams at supercomputing or hyperscaler infrastructure scale
• Led teams through rapid-growth periods where hiring and onboarding competed with roadmap delivery

About the role

We're looking for Research Engineers to build the evaluations that tell us — and the world — what Claude can actually do. Your work will turn ambiguous notions of "intelligence" into clear, defensible metrics that researchers, leadership, and the public can rely on.

You'll design and implement evaluations across the full spectrum of Claude's capabilities and personality, and build the infrastructure that runs them reliably at scale. You'll partner closely with researchers throughout the lifecycle of a new capability — from defining what to measure, to running the eval against live training checkpoints, to interpreting the results. The goal is to make Anthropic the leader in extremely well-characterized AI systems, with performance that is exhaustively measured and validated across the tasks that matter.

Key responsibilities

• Design and run new evaluations of Claude's capabilities — reasoning, agentic behavior, knowledge, safety properties — and produce visualizations that make the results legible to researchers and decision-makers
• Build and harden the distributed eval execution platform so hundreds of evals run reliably against checkpoints throughout production RL training runs
• Own the dashboards researchers and leadership use to monitor model health during training, improving signal-to-noise, reducing latency, and making regressions impossible to miss
• Debug anomalous eval results mid-training-run, determine whether the cause is a model change or an infrastructure issue, and communicate the answer clearly under time pressure
• Improve the tooling, libraries, and workflows researchers use to implement and iterate on evaluations
• Partner with research teams across the full lifecycle of a new capability — from defining what to measure to interpreting results as training progresses
• Run experiments to characterize how prompting, sampling, and scaffolding choices affect results on internal and industry benchmarks
• Communicate evaluations and their results to internal stakeholders and, where appropriate, external audiences

Minimum qualifications

• Strong Python programming skills, including production or research infrastructure
• Experience building or operating distributed systems, data pipelines, or other infrastructure that needs to be reliable at scale
• Clear written and verbal communication, especially when explaining technical results to non-specialists
• Comfort operating in an on-call or production-support capacity when training runs are live
• Care about the societal impacts of your work and an interest in steering powerful AI to be safe and beneficial

Preferred qualifications

• Hands-on experience using large language models such as Claude, including prompting, sampling, and scaffolding
• Background in data visualization and a track record of building dashboards people actually trust and use
• Experience developing robust evaluation metrics for language models
• Experience with observability, monitoring, or experiment-tracking systems
• Background in statistics and experimental design
• Experience with large-scale dataset sourcing, curation, and processing
• Experience running or supporting ML training infrastructure
• A bias toward picking up slack and operating flexibly across team boundaries
• Enjoy pair programming — we love to pair

Representative projects

• Stand up a new eval that tests a specific reasoning capability from scratch — define the task, build the dataset, implement the scoring, validate against known signals, and ship a dashboard that makes the result legible
• Diagnose a mid-training regression: an eval suite returns anomalous numbers, and you need to determine within hours whether it's the model, the harness, the data, or the infrastructure
• Take a flaky distributed eval pipeline and make it boring — better retries, better observability, faster feedback to researchers
• Partner with a research team on a new capability area, helping them articulate what "good" looks like and translating that into measurable artifacts

About the role

Anthropic is seeking a Senior Full Stack Engineer to help build the technical foundation for how Anthropic educates customers and enterprises at scale, and to shape what that even means when AI is the delivery mechanism, not just the subject being taught.

Your work will define how millions of people experience learning about Claude: how content is served and adapted in real time, how assessments respond to what a learner actually understands, how credentialing systems verify genuine competence rather than course completion, and how the entire platform evolves as fast as the product it teaches. You'll build the infrastructure that makes AI-native education real — not as a concept, but as a working system that people interact with every day.

You’ll own the technical vision for our AI-driven educational experience: the platform layer, enterprise readiness features, the data architecture, the AI-augmented pipelines, the integrations, and the delivery systems. You'll work with a handful of highly technical educational collaborators to own the technical vision and execution for how Anthropic’s education team delivers educational experiences at scale, and you'll build the AI-augmented pipelines and platform capabilities that let a small team of educators operate with reach and responsiveness that would otherwise require an organization ten times larger.

This is an engineering role for someone who genuinely cares about how people learn. You'll make platform decisions that are deeply educational, and you need the judgment to make those decisions well, not just implement someone else's spec. You'll interface closely with Anthropic's security, infrastructure, and data teams, and you'll partner with trainers and educators on the GTM, devrel, and education teams who will be your primary stakeholders.

Responsibilities

• Architect and own Anthropic’s enterprise ready education platform and infrastructure — authentication/SSO, learner data models, content delivery, credentialing, and progress tracking — designed for scale, AI-forward features, and reliability from day one

• Own the technical vision for how Anthropic delivers AI-driven educational experiences and shaping what's possible, not just building what's requested. You’ll determine what to build, what to extend, and what to replace as needs evolve

• Build AI-augmented pipelines for content generation, real-time adaptation, assessment delivery, and quality assurance — working closely with the education team to understand what these systems need to do pedagogically

• Design platform capabilities that enable genuinely new kinds of empowering learning experiences — adaptive paths, competency-based progression, AI-evaluated demonstrations of skill — not just digital versions of traditional courses

• Interface with Anthropic's security, infrastructure, and data teams to ensure education systems meet organizational standards and integrate cleanly with existing architecture

• Build platform abstractions that let non-engineers on the team configure, extend, and experiment with educational experiences without engineering bottlenecks

• Make product-level decisions about how the platform works — you're shaping what the learning experience is, not just serving it

You may be a good fit if you have

• 7+ years of software engineering experience, with demonstrated ability to take a platform from zero to production at scale

• Genuine passion for education — you think about how people learn, you have opinions about what educational technology should look like in an AI-native world, and you want to build it

• A strong eye for quality educational content, with a sense of the complex systems that produce it

• Full-stack engineering skills with particular strength in backend systems: databases, APIs, authentication, reliability, and infrastructure

• Experience building and operating platforms that serve many concurrent users reliably

• Comfort with ambiguity and ownership: you'll need to translate pedagogical intent into technical architecture without heavy specification

• Experience interfacing with security and infrastructure teams to meet organizational requirements while maintaining development velocity

• A working practice of using AI tools in your own engineering workflows — and genuine curiosity about how AI fundamentally changes what educational platforms can be

Strong candidates may also have

• Experience building or significantly contributing to learning platforms, credentialing systems, or educational technology products

• Background in or exposure to learning science, instructional design, or educational research

• Experience with SSO/identity (SAML, OAuth), badging standards (Open Badges), or LTI integrations

• Familiarity with building systems that incorporate LLM capabilities as core infrastructure

• Prior work at a high-growth company where you built platform infrastructure that other teams depended on

• Experience designing systems that non-technical users configure and operate

About the role

The RL Velocity team owns the efficiency and reliability of our RL Science stack - the infrastructure, tooling, and systems that let researchers iterate quickly on training runs. As a Research Engineer on the team, you'll build and improve the core platform that underpins how we do RL at Anthropic, removing bottlenecks that slow down research and making it easier for the broader org to ship better models faster. This is high-leverage work: small improvements to velocity compound across every researcher and every run.

Responsibilities

• Build and improve the RL training infrastructure that researchers depend on day-to-day
• Identify and remove bottlenecks across the RL stack: debugging, profiling, and rearchitecting where needed
• Partner closely with researchers and with adjacent engineering teams (inference, sandboxing, and many more) to understand pain points and ship tooling that makes them faster
• Own the reliability and performance of research runs end-to-end
• Contribute to design decisions that shape how Anthropic does RL at scale

You may be a good fit if you

• Have strong software engineering fundamentals and a track record of building performant, reliable systems
• Have worked on ML infrastructure, distributed systems, or research tooling
• Care about enabling other people's work and find leverage through platforms rather than individual experiments
• Are comfortable operating across the stack, from low-level performance work to RL algorithms
• Have a bias toward shipping and iterating quickly, with a mix of high agency and low ego

Strong candidates may also have

• Experience with large-scale distributed training (RL, pre-training, or post-training)
• Familiarity with JAX, PyTorch, or similar ML frameworks
• A track record of operating at the edge of research and infra in a fast-moving environment

Deadline to apply: None. Applications will be reviewed on a rolling basis.

About the role

The RL Velocity team owns the efficiency and reliability of our RL Science stack - the infrastructure, tooling, and systems that let researchers iterate quickly on training runs. As a Research Engineer on the team, you'll build and improve the core platform that underpins how we do RL at Anthropic, removing bottlenecks that slow down research and making it easier for the broader org to ship better models faster. This is high-leverage work: small improvements to velocity compound across every researcher and every run.

Responsibilities

• Build and improve the RL training infrastructure that researchers depend on day-to-day
• Identify and remove bottlenecks across the RL stack: debugging, profiling, and rearchitecting where needed
• Partner closely with researchers and with adjacent engineering teams (inference, sandboxing, and many more) to understand pain points and ship tooling that makes them faster
• Own the reliability and performance of research runs end-to-end
• Contribute to design decisions that shape how Anthropic does RL at scale

You may be a good fit if you

• Have strong software engineering fundamentals and a track record of building performant, reliable systems
• Have worked on ML infrastructure, distributed systems, or research tooling
• Care about enabling other people's work and find leverage through platforms rather than individual experiments
• Are comfortable operating across the stack, from low-level performance work to RL algorithms
• Have a bias toward shipping and iterating quickly, with a mix of high agency and low ego

Strong candidates may also have

• Experience with large-scale distributed training (RL, pre-training, or post-training)
• Familiarity with JAX, PyTorch, or similar ML frameworks
• A track record of operating at the edge of research and infra in a fast-moving environment

Deadline to apply: None. Applications will be reviewed on a rolling basis.

About the Team

Safeguards Labs is a new team operating at the intersection of research and engineering, chartered to investigate novel safety methods that protect Claude and the people who use it. We prototype new approaches to safe models, usage safeguards, and production safety — pressure-testing ideas through offline analysis and subsets of traffic before they graduate into production systems run by our partner Safeguards teams. Our work overlaps closely with account abuse, model behavior safeguards, and other safeguard subteams, and we serve as a research arm that can take on ambitious, ambiguous problems and turn them into deployed defenses.

About the Role

We're hiring research engineers to define and execute the Labs research agenda. You'll scope your own projects, run experiments end-to-end, and decide when an idea is ready to hand off to a production team — or when to kill it and move on. The team is small and being built deliberately around a roughly 3:1 mix of researchers to software engineers, so each person has substantial latitude over what they work on and high leverage on the team's direction.

Responsibilities:

• Lead and contribute to research projects investigating new methods for detecting misuse of Claude, identifying malicious organizations and accounts, strengthening model safeguards, and other safety needs.
• Design and run offline analyses over model usage data to surface abuse patterns, build classifiers and detection systems, and evaluate their effectiveness.
• Develop and iterate on prototypes that could eventually feed signals into the real-time safeguards path, partnering with engineers on tech transfer.
• Contribute to a broader research portfolio investigating methods for detecting abusive behavior in chat-based or agentive workflows, and for training the model to robustly refrain from dangerous responses or behaviors without over-refusing.
• Build evaluations and methodologies for measuring whether safeguards actually work, including in agentic settings.
• Write up findings clearly so they inform decisions across Trust & Safety, research, and product teams.

You may be a good fit if you:

• Have a track record of independently driving research projects from ambiguous problem statements to concrete results, ideally in AI, ML, security, integrity, or a related technical field.
• Are comfortable scoping your own work and switching between research, engineering, and analysis as a project demands.
• Have working familiarity with how large language models operate — sampling, prompting, training — even if LLMs aren't your primary background.
• Are proficient in Python and comfortable working with large datasets.
• Care about the societal impacts of AI and want your work to directly reduce real-world harm.

Strong candidates may also have: 

• Experience building and training machine learning models, including classifiers for abuse, fraud, integrity, or security applications.
• Knowledge of evaluation methodologies for language models and experience designing evals.
• Experience with agentic environments and evaluating model behavior in them.
• Background in trust and safety, integrity, fraud detection, threat intelligence, or adversarial ML.
• Experience with red teaming, jailbreak research, or interpretability methods like steering vectors.
• A history of taking research prototypes and transferring them into production systems.

Program Overview

We're working toward an AI scientist — a system capable of the long-horizon reasoning and experimental judgment needed to push the scientific frontier. We believe frontier AI, developed safely, can meaningfully accelerate progress across science, engineering, and beyond. Getting there requires close collaboration with people who have spent years working in these domains.

The Anthropic STEM Fellows Program brings students and experts from STEM domains into Anthropic for a few months to work alongside our research teams. Fellows pair deep domain expertise with our frontier models and internal tooling to evaluate, improve, and apply Claude's capabilities in their field.

We're looking for Fellows across all different science and engineering domains. Prior machine learning experience is helpful but not required — scientific judgment and a willingness to learn quickly matter more.

What to Expect:

• A concrete, scoped project matched to your background, with direct mentorship from Anthropic researchers

• Access to frontier Claude models and internal evaluation tooling

• A cohort of incoming Fellows to collaborate with and connection to Anthropic's broader research community

What You'll Work On:

• Designing rigorous, hard-to-game evaluations that measure whether models can plan experiments, interpret data, and reason about mechanisms in your field, and systematically finding where they're confidently wrong.

• Identifying or creating data and techniques that target capability gaps.

• Applying Claude to open problems in your domain, by using various test-time compute strategies and integrating relevant scientific tools/software.

Example projects might look like a materials scientist who finds Claude struggling to reason about phase stability, and builds a pipeline producing property prediction tasks to address this gap, or a climate scientist integrating atmospheric modeling tools and building environments that exercise them. Projects will be scoped to ship within the period of the fellowship.

Ideal Candidates:

• Are excited about the potential for AI to accelerate scientific discovery

• Are familiar with using Claude and Claude Code as part of their day-to-day workflow

• Are comfortable working in a fast-paced, high-agency atmosphere for a few months and delivering concrete results

• Hold (or are completing) a PhD in a STEM field, or have equivalent research experience.

Application Process

Candidates will need to progress through the following stages:

1. Submit an application

2. Complete an initial technical assessment

3. Complete a take-home exercise

4. Chat with a mentor about research and potential project directions

We will accept applications until May 15 and plan to have final decisions by June 1. 

Logistics:

• Duration: The program will start June 15 and last until September 15; some fellowships may be considered for an extension.

• Location: In-person at our San Francisco office.

• Commitment: Full-time role for the duration of the program.

• Compensation: Weekly stipend of ~$3,800, access to our offices, and other benefits.

• Authorization to work in the US required for the duration of the program.

The below are Anthropic's policies for full time roles. These do NOT apply to the Fellows Program.

About the role

The Research Data Platform team builds the tools that Anthropic's researchers use every day to manage, query, and analyze the data that goes into training and evaluating frontier models. We power the internal applications researchers rely on to monitor RL runs, explore finetuning datasets, and understand what's happening inside their experiments.

We're looking for engineers who love working directly with users and who excel at building data products — the pipelines that move data out of training runs into queryable storage, and the APIs, libraries, and services researchers use to manage and explore it. This role sits closer to the research workflow than a typical data infrastructure position: you'll often embed with research teams, build ML-specific tooling alongside them, and leverage what our Data Infrastructure team has already built rather than reinventing it.

We do not require prior ML or AI training experience. If you enjoy working closely with technical users, learning new domains quickly, and building tools people actually want to use, you'll pick up the research context fast.

Responsibilities

• Build and operate data pipelines that extract data from research training runs and land it in storage systems that are easy and fast to query
• Work closely with researchers to design and build APIs, libraries, and web interfaces that support data management, exploration, and analysis
• Develop dataset management, data cataloging, and provenance tooling that researchers use in their day-to-day work
• Embed with research teams to understand their workflows, identify high-leverage tooling opportunities, and ship solutions quickly
• Collaborate with adjacent teams to build on existing systems rather than reinventing them

You may be a good fit if you

• Have significant software engineering experience, particularly building data-intensive applications or internal tooling
• Enjoy working directly with users, gathering requirements iteratively, and shipping things that get adopted
• Are results-oriented, with a bias towards flexibility and impact
• Pick up slack, even if it goes outside your job description
• Want to learn more about machine learning research
• Care about the societal impacts of your work

Strong candidates may also have experience with

• Large-scale ETL, columnar storage formats, and query engines (e.g., Spark, BigQuery, DuckDB, Parquet)
• High-volume time series data — ingestion, storage, and efficient querying
• Data cataloging, lineage, or metadata management systems
• ML experiment tracking or metrics platforms
• Working in environments where engineers partner closely with quantitative users — research labs, trading firms, observability or analytics startups
• Complex data visualization and full-stack web application development

About the Team

As AI training and deployments scale, the volume of data we need to monitor and understand is exploding. Our team uses Claude itself to make sense of this data. We own an integrated set of tools enabling Anthropic to ask open-ended questions, surface unexpected patterns, and maintain meaningful human oversight over massive datasets.

Our tools are widely adopted internally — powering ongoing enforcement, threat intelligence investigations, model audits, and more — and we’re looking for experienced engineers and researchers to both scale up existing applications and go zero-to-one on new ones.

About the Role

As a Research Engineer on our team, you'll design and build systems that let AI analyze large, unstructured datasets — think tens or hundreds of thousands of conversations or documents — and produce structured, trustworthy insights. You'll work across the full stack, from core analysis frameworks through user-facing apps and interfaces.

This is a high-leverage role. The tools you build will be used by dozens of researchers and investigators, and directly shape our ability to measure and mitigate both misuse and misalignment.

 Responsibilities:

• Design and implement AI-based monitoring systems for AI training and deployment
• Extend and improve core frameworks for processing large volumes of unstructured text
• Partner with researchers and safety teams across Anthropic to understand their analytical needs and build solutions
• Develop agentic integrations that allow AI systems to autonomously investigate and act on analytical findings
• Contribute to the strategic direction of the team, including decisions about what to build, what to partner on, and where to invest

You May Be a Good Fit If You:

• Have 5+ years of software engineering experience, with meaningful exposure to ML systems
• Are excited about the problem of scaling human oversight of AI systems
• Are familiar with LLM application development and evaluation 
• Enjoy building tools that other people use — you care about UX, reliability, and documentation
• Thrive in collaborative, cross-functional environments

Strong Candidates May Also Have:

• Experience with productionizing internal tools or building developer-facing platforms
• Background in building monitoring or observability systems
• Comfort with ambiguity — our team is small and growing, and you'll help define what we become

About the Team

Our team is organized around the north star goal of building an AI scientist—a system capable of solving the long-term reasoning challenges and basic capabilities necessary to push the scientific frontier.

About the Role

We're seeking a Science Research Operations team member to build and own the operational infrastructure that keeps our research organization running at full speed. Our science teams are working on some of the hardest and most consequential problems in AI—training large-scale models, running complex experiments, and building novel products at the frontier. What makes that possible isn't just talent: it's the coordination, systems, and programs that let researchers spend their time on the science rather than the overhead around it.

This role sits at the intersection of research operations, technical program management, and product strategy. You'll work directly with research scientists and research engineers, doing a mix of tasks including running research partnerships, managing complex internal programs, and helping run the team’s day-to-day operations. You'll also contribute to science product development—helping translate research directions into product strategy and ensuring our production deployment environments reflect our best configurations.

This is not a pure coordination role. The best candidates will engage substantively with what the team is building, have a role in determining our strategy, spot problems before they surface, and bring genuine ownership to the systems and programs they run.

Responsibilities:

• Build and manage custom expert contractor networks, sourcing domain specialists for eval and training data work that requires expertise beyond standard channels
• Run research partnerships with external partners, from scoping through delivery
• Provide end-to-end TPM support for major research pushes—coordinating across teams, tracking dependencies, and keeping stakeholders aligned
• Ensure that our research progress is complemented by products that enable scientists to make maximal use of model capabilities. 
• Support recruiting efforts. 
• Coordinate external communications for the team, including supporting blog posts and preparing public-facing materials
• Partner with product teams to contribute to science product strategy, product design, and novel product integrations where research and product intersect
• Own team logistics including onboarding coordination, team events, and operational programs that improve team efficiency

You may be a good fit if you:

• Have experience in research operations, technical program management, or a related role in a fast-moving technical environment
• Can context-switch fluidly between operational work (logistics, tracking, coordination) and higher-order work (strategy, partnerships, product thinking)
• Have a technical background, with experience in software development, machine learning, or biology R&D. 
• Are comfortable working directly with research scientists and engineers—you ask good questions, you don't need things explained twice, and you know when to escalate vs. when to handle it yourself
• Have a track record of building systems and processes from scratch rather than inheriting them
• Bring strong written communication skills and can represent the team accurately in external-facing materials
• Have managed contractors or external partners before, including scoping work, tracking delivery, and ensuring quality
• Are results-oriented, with a bias toward flexibility and impact
• Thrive in ambiguous, fast-moving environments where priorities shift and no two weeks look the same

Strong candidates may also have:

• Direct experience sourcing and managing expert contractor networks, particularly in technical or scientific domains
• Familiarity with ML research workflows—training runs, evaluations, data pipelines—and what makes them succeed or stall
• Experience contributing to product development or product strategy, not just operations

This role offers a rare opportunity to have broad impact on a research organization working at the frontier of AI. You'll be a key part of how ambitious science gets done—not just facilitating it, but shaping the programs, products, and infrastructure that make it possible. If you're energized by high-ownership work at the intersection of operations and research, we'd love to hear from you.

Deadline to apply: None. Applications will be reviewed on a rolling basis.

About the team

We are seeking passionate Research Scientists and Engineers to join our growing Pre-training team in Zurich. We are involved in developing the next generation of large language models. The team primarily focuses on multimodal capabilities: giving LLMs the ability to understand and interact with modalities other than text.

In this role, you will work at the intersection of cutting-edge research and practical engineering, contributing to the development of safe, steerable, and trustworthy AI systems.

Responsibilities

In this role you will interact with many parts of the engineering and research stacks.

• Conduct research and implement solutions in areas such as model architecture, algorithms, data processing, and optimizer development

• Independently lead small research projects while collaborating with team members on larger initiatives

• Design, run, and analyze scientific experiments to advance our understanding of large language models

• Optimize and scale our training infrastructure to improve efficiency and reliability

• Develop and improve dev tooling to enhance team productivity

• Contribute to the entire stack, from low-level optimizations to high-level model design

Qualifications & Experience

We encourage you to apply even if you do not believe you meet every single criterion. Because we focus on so many areas, the team is looking for both experienced engineers and strong researchers, and encourage anyone along the researcher/engineer spectrum to apply.

• Degree (BA required, MS or PhD preferred) in Computer Science, Machine Learning, or a related field

• Strong software engineering skills with a proven track record of building complex systems

• Expertise in Python and deep learning frameworks

• Have worked on high-performance, large-scale ML systems, particularly in the context of language modeling

• Familiarity with ML Accelerators, Kubernetes, and large-scale data processing

• Strong problem-solving skills and a results-oriented mindset

• Excellent communication skills and ability to work in a collaborative environment

You'll thrive in this role if you

• Have significant software engineering experience

• Are able to balance research goals with practical engineering constraints

• Are happy to take on tasks outside your job description to support the team

• Enjoy pair programming and collaborative work

• Are eager to learn more about machine learning research

• Are enthusiastic to work at an organization that functions as a single, cohesive team pursuing large-scale AI research projects

• Have ambitious goals for AI safety and general progress in the next few years, and you’re excited to create the best outcomes over the long-term

Sample Projects

• Optimizing the throughput of novel attention mechanisms

• Proposing Transformer variants, and experimentally comparing their performance

• Preparing large-scale datasets for model consumption

• Scaling distributed training jobs to thousands of accelerators

• Designing fault tolerance strategies for training infrastructure

• Creating interactive visualizations of model internals, such as attention patterns

If you're excited about pushing the boundaries of AI while prioritizing safety and ethics, we want to hear from you!

About the Role

As a Full-Stack Software Engineer in RL, you'll build the platforms, tools, and interfaces that power environment creation, data collection, and training observability. The quality of Claude's next generation depends on the quality of the data we train it on — and the systems you build are what make that data possible.

You'll own product surfaces end-to-end — from backend services and APIs to the web UIs that researchers, external vendors, and thousands of data labelers use every day. You don't need a background in ML research. What matters is that you can take an ambiguous, high-stakes problem and ship a polished, reliable product against it, fast.

This team moves very quickly. Claude writes a lot of the code we commit, which means the bottleneck isn't typing — it's judgment, taste, and the ability to react to what researchers need next. You'll iterate on data collection strategies to distill the knowledge of thousands of human experts around the world into our models, and you'll do it in a loop that closes in hours and days, not quarters or months.

Anthropic's Reinforcement Learning organization leads the research and development that trains Claude to be capable, reliable, and safe. We've contributed to every Claude model, with significant impact on the autonomy and coding capabilities of our most advanced models. Our work spans teaching models to use computers effectively, advancing code generation through RL, pioneering fundamental RL research for large language models, and building the scalable training methodologies behind our frontier production models.

The RL org is organized around four goals: solving the science of long-horizon tasks and continual learning, scaling RL data and environments to be comprehensive and diverse, automating software engineering end-to-end, and training the frontier production model. Our engineering teams build the environments, evaluation systems, data pipelines, and tooling that make all of this possible — from realistic agentic training environments and scalable code data generation to human data collection platforms and production training operations.

What You'll Do

• Build and extend web platforms for RL environment creation, management, and quality review — including environment configuration, versioning, and validation workflows
• Develop vendor-facing interfaces and tooling that let external partners create, submit, and iterate on training environments with minimal friction
• Design and implement platforms for human data collection at scale, including labeling workflows, quality assurance systems, and feedback mechanisms that surface reward signal integrity issues early
• Build evaluation dashboards and observability UIs that give researchers real-time insight into environment quality, training run health, and reward hacking
• Create backend services and APIs that connect environment authoring tools, data collection systems, and RL training infrastructure
• Build and expand scalable code data generation pipelines, producing diverse programming tasks with robust reward signals across languages and difficulty levels
• Develop onboarding automation and documentation tooling so new vendors and internal users ramp up in hours, not weeks
• Partner closely with RL researchers, data operations, and vendor management to translate ambiguous requirements into well-scoped, well-designed products

You May Be a Good Fit If You

• Have strong software engineering fundamentals and real full-stack range — you're comfortable owning a surface from database schema to frontend
• Are proficient in Python and a modern web stack (React, TypeScript, or similar)
• Have a track record of shipping systems that solved a hard problem, not just shipped on time — e.g. you built the thing that made your team 10x faster, or the internal tool nobody thought was possible
• Operate with high agency: you identify what needs to be done and drive it forward without waiting for a ticket
• Have found yourself wondering "why isn't this moving faster?" in previous roles — and then have done something about it
• Care about UX and can build interfaces that are intuitive for both technical researchers and non-technical labelers
• Communicate clearly with researchers, operations teams, and engineers, and can turn vague asks into well-scoped work
• Thrive in a fast-moving environment where priorities shift, Claude is your pair programmer, and the next problem is often one nobody has solved before
• Care about Anthropic's mission to build safe, beneficial AI and want your work to contribute directly to it

Strong Candidates May Also Have

• Built data collection, labeling, or annotation platforms — ideally ones that had to scale across many vendors or many task types
• Background building multi-tenant platforms with role-based access, audit trails, and vendor management workflows
• Experience with cloud infrastructure (GCP or AWS), Docker, and CI/CD pipelines
• Familiarity with LLM training, fine-tuning, or evaluation workflows
• Experience with async Python (Trio, asyncio) or high-throughput API design
• Background in dashboards, monitoring, or observability tooling
• Experience working directly with external vendors or partners on technical integrations
• A background that isn't a straight line — e.g. math or physics into SWE, competitive programming, research into engineering, or a side project that outgrew its scope

Representative Projects

• Building a unified platform for human data collection that integrates labeling workflows, vendor management, and QA for complex agentic tasks
• Developing vendor onboarding automation that handles Docker registry access, API token management, and environment validation
• Creating evaluation and observability dashboards that catch reward hacks, measure environment difficulty, and give real-time feedback during production training
• Building environment quality review workflows that let researchers browse, grade, and provide feedback on training environments
• Developing automated environment quality pipelines that validate correctness and difficulty calibration before environments hit production training
• Building internal tools for browsing and analyzing training run results, environment statistics, and data collection progress

About the Role

We are looking for researchers to help mitigate the risks that come with building AI systems. One of these risks is the potential for models to interact with private user data. In this role, you'll design and implement privacy-preserving techniques, audit our current techniques, and set the direction for how Anthropic handles privacy more broadly.

Responsibilities: 

• Lead our privacy analysis of frontier models, carefully auditing the use of data and ensuring safety throughout the process
• Develop privacy-first training algorithms and techniques
• Develop evaluation and auditing techniques to measure the privacy of training algorithms
• Work with a small, senior team of engineers and researchers to enact a forward-looking privacy policy
• Advocate on behalf of our users to ensure responsible handling of all data

You may be a good fit if you have:

• Experience working on privacy-preserving machine learning
• A track record of shipping products and features inside a fast-moving environment
• Strong coding skills in Python and familiarity with ML frameworks like PyTorch or JAX.
• Deep familiarity with large language models, how they work, and how they are trained
• Have experience working with privacy-preserving techniques (e.g., differential privacy and how it is different from k-anonymity, l-diversity, and t-closeness)
• Experience supporting fast-paced startup engineering teams
• Demonstrated success in bringing clarity and ownership to ambiguous technical problems
• Proven ability to lead cross-functional security initiatives and navigate complex organizational dynamics

Strong candidates may also:

• Have published papers on the topic of privacy-preserving ML at top academic venues
• Prior experience training large language models (e.g., collecting training datasets, pre-training models, post-training models via fine-tuning and RL, running evaluations on trained models)
• Prior experience developing tooling to support privacy-preserving ML (e.g., differential privacy in TF-Privacy or Opacus)

Anthropic is at the forefront of AI research, dedicated to developing safe, ethical, and powerful artificial intelligence. Our mission is to ensure that transformative AI systems are aligned with human interests. We are seeking a Research Engineer to join our Pre-training team, responsible for developing the next generation of large language models. In this role, you will work at the intersection of cutting-edge research and practical engineering, contributing to the development of safe, steerable, and trustworthy AI systems.

Key Responsibilities:

• Conduct research and implement solutions in areas such as model architecture, algorithms, data processing, and optimizer development
• Independently lead small research projects while collaborating with team members on larger initiatives
• Design, run, and analyze scientific experiments to advance our understanding of large language models
• Optimize and scale our training infrastructure to improve efficiency and reliability
• Develop and improve dev tooling to enhance team productivity
• Contribute to the entire stack, from low-level optimizations to high-level model design

Qualifications:

• Advanced degree (MS or PhD) in Computer Science, Machine Learning, or a related field
• Strong software engineering skills with a proven track record of building complex systems
• Expertise in Python and experience with deep learning frameworks (PyTorch preferred)
• Familiarity with large-scale machine learning, particularly in the context of language models
• Ability to balance research goals with practical engineering constraints
• Strong problem-solving skills and a results-oriented mindset
• Excellent communication skills and ability to work in a collaborative environment
• Care about the societal impacts of your work

Preferred Experience:

• Work on high-performance, large-scale ML systems
• Familiarity with GPUs, Kubernetes, and OS internals
• Experience with language modeling using transformer architectures
• Knowledge of reinforcement learning techniques
• Background in large-scale ETL processes

You'll thrive in this role if you:

• Have significant software engineering experience
• Are results-oriented with a bias towards flexibility and impact
• Willingly take on tasks outside your job description to support the team
• Enjoy pair programming and collaborative work
• Are eager to learn more about machine learning research
• Are enthusiastic to work at an organization that functions as a single, cohesive team pursuing large-scale AI research projects
• Are working to align state of the art models with human values and preferences, understand and interpret deep neural networks, or develop new models to support these areas of research
• View research and engineering as two sides of the same coin, and seek to understand all aspects of our research program as well as possible, to maximize the impact of your insights
• Have ambitious goals for AI safety and general progress in the next few years, and you’re working to create the best outcomes over the long-term.

Sample Projects:

• Optimizing the throughput of novel attention mechanisms
• Comparing compute efficiency of different Transformer variants
• Preparing large-scale datasets for efficient model consumption
• Scaling distributed training jobs to thousands of GPUs
• Designing fault tolerance strategies for our training infrastructure
• Creating interactive visualizations of model internals, such as attention patterns

At Anthropic, we are committed to fostering a diverse and inclusive workplace. We strongly encourage applications from candidates of all backgrounds, including those from underrepresented groups in tech.

If you're excited about pushing the boundaries of AI while prioritizing safety and ethics, we want to hear from you!

About the Team

The Frontier Red Team (FRT) is a small, focused technical research team within Anthropic's Policy organization. Our goal is to make the entire world safer in an era of advanced AI by understanding what these systems can do and building the defenses that matter.

In 2026, we're focused on researching and ensuring safety with self-improving, highly autonomous AI systems, especially ones related to cyberphysical capabilities. See our previous related work on exploits, partnering with Mozilla, and zero days. This is early-stage, high-conviction research with the potential for outsized impact — Glasswing is one example.

Note: We are exclusively hiring in SF. We support relocation, but all hires must relocate before starting.

About the Role

In the last year, we've seen compelling signs that LLMs and agents are increasingly capable of novel cyber capabilities. We think 2026 will be the year where models reach expert-level, even superhuman, in several cybersecurity domains. This is a novel and massive threat surface.

As a Research Scientist on FRT focusing on cyber, you'll build the tools and frameworks needed to defend the world against advanced AI-enabled cyber threats. Senior candidates will have the opportunity to shape and grow Anthropic's cyberdefense research program, working with Security, Safeguards, Policy, and other partner teams. This work sits at the intersection of AI capabilities research, cybersecurity, and policy—what we learn directly shapes how Anthropic and the world prepare for AI-enabled cyber threats.

This is applied research with real-world stakes. Your work will inform decisions at the highest levels of the company, contribute to demonstrations that shape policy discourse, and build the technical defenses that we will need for a future of increasingly powerful AI systems.

What You'll Do

• Develop systems, tools, and frameworks for AI-empowered cybersecurity, such as autonomous vulnerability discovery and remediation, malware detection and management, network hardening, and pentesting
• Design and run experiments to elicit and evaluate autonomous AI cyber capabilities in realistic scenarios
• Design and build infrastructure for evaluating and enabling AI systems to operate in security environments
• Translate technical findings into compelling demonstrations and artifacts that inform policymakers and the public
• Collaborate with external experts in cybersecurity, national security, and AI safety to scope and validate research directions
• Senior candidates will also set research strategy, define what problems are worth solving, own the technical roadmap, and manage relationships with cross-functional partners

Sample Projects

• Building frameworks and tools that enable AI models to autonomously find and patch vulnerabilities
• Running purple-team simulations where AI defenders compete against AI attackers in network environments
• Pointing autonomous AI systems at real-world security challenges (bug bounties, CTFs etc.) to characterize risks, defensive potential, and compare to human experts
• Building demonstrations of frontier AI cyber capabilities for policy stakeholders

You May Be a Good Fit If You

• Have deep expertise in cybersecurity or security research
• Are driven to find solutions to complex, high-stakes problems
• Have experience doing technical research with LLM-based agents or autonomous systems
• Have strong software engineering skills, particularly in Python
• Can own entire problems end-to-end, including both technical and non-technical components
• Design and run experiments quickly, iterating fast toward useful results
• Thrive in collaborative environments
• Care deeply about AI safety and want your work to have real-world impact on how humanity navigates advanced AI
• Are comfortable working on sensitive projects that require discretion and integrity
• Have proven ability to lead cross-functional security initiatives and navigate complex organizational dynamics

Strong Candidates May Also Have

• Experience with offensive security research, vulnerability research, or exploit development
• Research or professional experience applying LLMs to security problems
• Track record in competitive CTFs, bug bounties, or other security-related competitions
• Experience building security tools or automation
• Track record of building demos or prototypes that communicate complex technical ideas
• Experience working with external stakeholders (policymakers, government, researchers)
• Familiarity with AI safety research and threat modeling for advanced AI systems

About the Role

About the RL Teams

Our Reinforcement Learning teams lead Anthropic's reinforcement learning research and development, playing a critical role in advancing our AI systems. We've contributed to all Claude models, with significant impacts on the autonomy and coding capabilities of Claude Sonnet 4.6 and Opus 4.6. Our work spans several key areas:

• Developing systems that enable models to use computers effectively

• Advancing code generation through reinforcement learning

• Pioneering fundamental RL research for large language models

• Building scalable RL infrastructure and training methodologies

• Enhancing model reasoning capabilities

We collaborate closely with Anthropic's alignment and frontier red teams to ensure our systems are both capable and safe. We partner with the applied production training team to bring research innovations into deployed models, and are dedicated to implement our research at scale. Our Reinforcement Learning teams sit at the intersection of cutting-edge research and engineering excellence, with a deep commitment to building high-quality, scalable systems that push the boundaries of what AI can accomplish.

About the Role

We're hiring for the Code RL team within the RL organization. As a Research Engineer, you'll advance our models' ability to safely write correct, fast code for accelerators.

You'll need to know accelerator performance well to turn it into tasks and signals models can learn from. Specifically, you will:

• Invent, design and implement RL environments and evaluations.

• Conduct experiments and shape our research roadmap.

• Deliver your work into training runs.

• Collaborate with other researchers, engineers, and performance engineering specialists across and outside Anthropic.

You may be a good fit if you:

• Have expertise with accelerators (CUDA, ROCm, Triton, Pallas), ML framework programming (JAX or PyTorch).

• Have worked across the stack – kernels, model code, distributed systems.

• Know how to balance research exploration with engineering implementation.

• Are passionate about AI's potential and committed to developing safe and beneficial systems.

Strong candidates may also have:

• Experience with reinforcement learning.

• Experience porting ML workloads between different types of accelerators.

• Familiarity with LLM training methodologies.

About the Team

Our team is organized around the north star goal of building an AI scientist – a system capable of solving the long term reasoning challenges and basic capabilities necessary to push the scientific frontier.

About the role

As a Research Engineer on our team you will work end to end across the whole model stack, identifying and addressing key infra blockers on the path to scientific AGI. Strong candidates should have familiarity with elements of language model training, evaluation, and inference and eagerness to quickly dive and get up to speed in areas they are not yet an expert on. This may include performance optimization, distributed systems, VM/sandboxing/container deployment, and large scale data pipelines. Join us in our mission to develop advanced AI systems pushing the frontiers of science and benefiting humanity.

Responsibilities:

• Design and implement large-scale infrastructure systems to support AI scientist training, evaluation, and deployment across distributed environments
• Identify and resolve infrastructure bottlenecks impeding progress toward scientific capabilities
• Develop robust and reliable evaluation frameworks for measuring progress towards scientific AGI.
• Build scalable and performant VM/sandboxing/container architectures to safely execute long-horizon AI tasks and scientific workflows
• Collaborate to translate experimental requirements into production-ready infrastructure
• Develop large scale data pipelines to handle advanced language model training requirements
• Optimize large scale training and inference pipelines for stable and efficient reinforcement learning

You may be a good fit if you:

• Have 6+ years of highly-relevant experience in infrastructure engineering with demonstrated expertise in large-scale distributed systems
• Are a strong communicator and enjoy working collaboratively
• Possess deep knowledge of performance optimization techniques and system architectures for high-throughput ML workloads
• Have experience with containerization technologies (Docker, Kubernetes) and orchestration at scale
• Have proven track record of building large-scale data pipelines and distributed storage systems
• Excel at diagnosing and resolving complex infrastructure challenges in production environments
• Can work effectively across the full ML stack from data pipelines to performance optimization
• Have experience collaborating with other researchers to scale experimental ideas
• Thrive in fast-paced environments and can rapidly iterate from experimentation to production

Strong candidates may also have:

• Experience with language model training infrastructure and distributed ML frameworks (PyTorch, JAX, etc.)
• Background in building infrastructure for AI research labs or large-scale ML organizations
• Knowledge of GPU/TPU architectures and  language model inference optimization
• Experience with cloud platforms (AWS, GCP) at enterprise scale
• Familiarity with VM and container orchestration.
• Experience with workflow orchestration tools and experiment management systems
• History working with large scale reinforcement learning
• Comfort with large scale data pipelines (Beam, Spark, Dask, …)

About the role

We are looking for ML Engineers and Research Engineers to help detect and mitigate misuse of our AI systems. As a member of the Safeguards ML team, you will build systems that identify harmful use—from individual policy violations to sophisticated, coordinated attacks—and develop defenses that keep our products safe as capabilities advance. You will also work on systems that protect user wellbeing and ensure our models behave appropriately across a wide range of contexts. This work feeds directly into Anthropic's Responsible Scaling Policy commitments.

Responsibilities

• Develop classifiers to detect misuse and anomalous behavior at scale. This includes developing synthetic data pipelines for training classifiers and methods to automatically source representative evaluations to iterate on
• Build systems to monitor for harms that span multiple exchanges, such as coordinated cyber attacks and influence operations, and develop new methods for aggregating and analyzing signals across contexts
• Evaluate and improve the safety of agentic products—developing both threat models and environments to test for agentic risks, and developing and deploying mitigations for prompt injection attacks
• Conduct research on automated red-teaming, adversarial robustness, and other research that helps test for or find misuse

You may be a good fit if you

• Have 4+ years of experience in ML engineering, research engineering, or applied research, in academia or industry
• Have proficiency in Python and experience building ML systems
• Are comfortable working across the research-to-deployment pipeline, from exploratory experiments to production systems
• Are worried about misuse risks of AI systems, and want to work to mitigate them
• Have strong communication skills and ability to explain complex technical concepts to non-technical stakeholders

Strong candidates may also have experience with

• Language modeling and transformers
• Building classifiers, anomaly detection systems, or behavioral ML
• Adversarial machine learning or red-teaming
• Interpretability or probes
• Reinforcement learning
• High-performance, large-scale ML systems

About the role

Anthropic's production models undergo sophisticated post-training processes to enhance their capabilities, alignment, and safety. As a Research Engineer on our Post-Training team, you'll train our base models through the complete post-training stack to deliver the production Claude models that users interact with.

You'll work at the intersection of cutting-edge research and production engineering, implementing, scaling, and improving post-training techniques like Constitutional AI, RLHF, and other alignment methodologies. Your work will directly impact the quality, safety, and capabilities of our production models.

Note: For this role, we conduct all interviews in Python. This role may require responding to incidents on short-notice, including on weekends.

Responsibilities:

• Implement and optimize post-training techniques at scale on frontier models
• Conduct research to develop and optimize post-training recipes that directly improve production model quality
• Design, build, and run robust, efficient pipelines for model fine-tuning and evaluation
• Develop tools to measure and improve model performance across various dimensions
• Collaborate with research teams to translate emerging techniques into production-ready implementations
• Debug complex issues in training pipelines and model behavior
• Help establish best practices for reliable, reproducible model post-training

You may be a good fit if you:

• Thrive in controlled chaos and are energised, rather than overwhelmed, when juggling multiple urgent priorities
• Adapt quickly to changing priorities
• Maintain clarity when debugging complex, time-sensitive issues
• Have strong software engineering skills with experience building complex ML systems
• Are comfortable working with large-scale distributed systems and high-performance computing
• Have experience with training, fine-tuning, or evaluating large language models
• Can balance research exploration with engineering rigor and operational reliability
• Are adept at analyzing and debugging model training processes
• Enjoy collaborating across research and engineering disciplines
• Can navigate ambiguity and make progress in fast-moving research environments

Strong candidates may also:

• Have experience with LLMs
• Have a keen interest in AI safety and responsible deployment

We welcome candidates at various experience levels, with a preference for senior engineers who have hands-on experience with frontier AI systems. However, proficiency in Python, deep learning frameworks, and distributed computing is required for this role.

About the role:

You want to build and run elegant and thorough machine learning experiments to help us understand and steer the behavior of powerful AI systems. You care about making AI helpful, honest, and harmless, and are interested in the ways that this could be challenging in the context of human-level capabilities. You could describe yourself as both a scientist and an engineer. As a Research Engineer on Alignment Science, you'll contribute to exploratory experimental research on AI safety, with a focus on risks from powerful future systems (like those we would designate as ASL-3 or ASL-4 under our Responsible Scaling Policy), often in collaboration with other teams including Interpretability, Fine-Tuning, and the Frontier Red Team.

Our blog provides an overview of topics that the Alignment Science team is either currently exploring or has previously explored. For the London team, we are opportunistically hiring for the following research areas:

• AI Control: Creating methods to ensure advanced AI systems remain safe and harmless in unfamiliar or adversarial scenarios.
• Alignment Stress-testing: Creating model organisms of misalignment to improve our empirical understanding of how alignment failures might arise.

Note: Currently, the team's hub is in San Francisco, so we require all candidates to be based at least 25% in London and travel to San Francisco occasionally. Additionally, we are prioritizing growing our San Francisco teams, so you may not hear back on your application to the London team unless we see an unusually strong fit. For this role, we conduct all interviews in Python.

Representative Projects:

About the role

We are looking for a Research Engineer to help us train Claude specifically for virtual collaborator workflows. While Claude excels at general tasks, a lot of knowledge work requires targeted training on real organizational data and workflows. Your job will be to design and implement reinforcement learning (RL) environments that transform Claude into the best virtual collaborator, training on realistic tasks from navigating internal knowledge to creating financial models.

Responsibilities:

• Training Claude on document manipulation with good taste, including understanding, enhancing, and co-creating (e.g., Office doc formats, data visualization)
• Designing and implementing reinforcement learning pipelines targeted at virtual collaborator use cases (productivity, organizational navigation, vertical domains)
• Building and scaling our data creation platform for generating high-quality, open-ended tasks with domain experts and crowdworkers Integrating real organizational data to create realistic training environments
• Developing robust evaluation systems that maintain quality while avoiding reward hacking
• Partnering directly with product teams (e.g., Cowork, claude.ai) to ensure training aligns with product features

You may be a good fit if you:

• Are a very experienced Python programmer who can quickly produce reliable, high quality code that your teammates love using
• Have 5-8 years of strong machine learning experience
• Thrive at the intersection of research and product, with a pragmatic approach to solving real-world problems
• Are comfortable with ambiguity and can balance research rigor with shipping deadlines
• Enjoy collaborating across multiple teams (data operations, model training, product)
• Can context-switch between research problems and product engineering tasks
• Care about making AI genuinely helpful for everyday enterprise workflows

Strong candidates may also have experience with:

• Creating RL envs for realistic tasks.
• Reward modeling and preventing reward hacking
• Building human-in-the-loop training systems or crowdsourcing platforms
• Working with enterprise tools and APIs (Google Workspace, Microsoft Office, Slack, etc.)
• Developing evaluation frameworks for open-ended tasks
• Domain expertise in finance, legal, or healthcare workflows 
• Creating scalable data pipelines with quality control mechanisms
• Translating product requirements into technical training objectives 

Deadline to apply: None. Applications will be reviewed on a rolling basis. 

About the role: 

As a Research Scientist/Engineer focused on honesty within the Finetuning Alignment team, you'll spearhead the development of techniques to minimize hallucinations and enhance truthfulness in language models. Your work will focus on creating robust systems that are accurate and reflect their true levels of confidence across all domains, and that work to avoid being deceptive or misleading. Your work will be critical for ensuring our models maintain high standards of accuracy and honesty across diverse domains.

Note: The team is based in New York and so we have a preference for candidates who can be based in New York. For this role, we conduct all interviews in Python. We have filled our headcount for 2025. However, we are leaving this form open as an expression of interest since we expect to be growing the team in the future, and we will review your application when we do. As such, you may not hear back on your application to this team until the new year

Responsibilities:

• Design and implement novel data curation pipelines to identify, verify, and filter training data for accuracy given the model’s knowledge
• Develop specialized classifiers to detect potential hallucinations or miscalibrated claims made by the model
• Create and maintain comprehensive honesty benchmarks and evaluation frameworks
• Implement techniques to ground model outputs in verified information, such as search and retrieval-augmented generation (RAG) systems
• Design and deploy human feedback collection specifically for identifying and correcting miscalibrated responses
• Design and implement prompting pipelines to generate data that improves model accuracy and honesty
• Develop and test novel RL environments that reward truthful outputs and penalize fabricated claims
• Create tools to help human evaluators efficiently assess model outputs for accuracy

You may be a good fit if you:

• Have an MS/PhD in Computer Science, ML, or related field
• Possess strong programming skills in Python
• Have industry experience with language model finetuning and classifier training
• Show proficiency in experimental design and statistical analysis for measuring improvements in calibration and accuracy
• Care about AI safety and the accuracy and honesty of both current and future AI systems
• Have experience in data science or the creation and curation of datasets for finetuning LLMs
• An understanding of various metrics of uncertainty, calibration, and truthfulness in model outputs

Strong candidates may also have:

• Published work on hallucination prevention, factual grounding, or knowledge integration in language models
• Experience with fact-grounding techniques
• Background in developing confidence estimation or calibration methods for ML models
• A track record of creating and maintaining factual knowledge bases
• Familiarity with RLHF specifically applied to improving model truthfulness
• Worked with crowd-sourcing platforms and human feedback collection systems
• Experience developing evaluations of model accuracy or hallucinations

Join us in our mission to ensure advanced AI systems behave reliably and ethically while staying aligned with human values. 

About the role

We are looking for software engineers to help build safety and oversight mechanisms for our AI systems. As a software engineer on the Safeguards team, you will work to monitor models, prevent misuse, and ensure user well-being. This role will focus on building systems to detect unwanted model behaviors and prevent disallowed use of models. You will apply your technical skills to uphold our principles of safety, transparency, and oversight while enforcing our terms of service and acceptable use policies.

Responsibilities:

• Develop monitoring systems to detect unwanted behaviors from our API partners and potentially take automated enforcement actions; surface these in internal dashboards to analysts for manual review
• Build abuse detection mechanisms and infrastructure 
• Surface abuse patterns to our research teams to harden models at the training stage
• Build robust and reliable multi-layered defenses for real-time improvement of safety mechanisms that work at scale

You may be a good fit if you:

• Bachelor’s degree in Computer Science, Software Engineering or comparable experience
• 5-10+ years of experience in a software engineering position, preferably with a focus on integrity, spam, fraud, or abuse detection and mitigation
• Proficiency in Python and Typescript
• Ability to work across the stack
• Strong communication skills and ability to explain complex technical concepts to non-technical stakeholders

Strong candidates may also:

• Have experience building trust and safety detection mechanisms and intervention for AI/ML systems
• Have experience with prompt engineering, jailbreak attacks, and other adversarial inputs
• Have worked closely with operational teams to build custom internal tooling

Deadline to apply: None. Applications will be reviewed on a rolling basis. 

About the role

As a Research Lead on the Training Insights team, you'll develop the strategy for, and lead execution on, how we measure and characterize model capabilities across training and deployment. This is a hands-on leadership role: you'll drive original research into new evaluation methodologies while leading a small team of researchers and research engineers doing the same.

Your work will span the full lifecycle of model development. You'll research and build new long-horizon evaluations that test the boundaries of what our models can achieve, develop novel approaches to measuring emerging capabilities, and deepen our understanding of how those capabilities develop — both during production RL training and after. You'll also take a cross-organizational view, working across Reinforcement Learning, Pretraining, Inference, Product, Alignment, Safeguards, and other teams to map the landscape of model evaluations at Anthropic and identify critical gaps in coverage.

This role carries significant visibility and impact. You'll help shape the evaluation narrative for model releases, contributing directly to how Anthropic communicates about its models to both internal and external audiences. Done well, you will change how the industry measures and understands model capabilities, significantly furthering our safety mission.  

Responsibilities: 

• Build new novel and long-horizon evaluations
• Develop novel measurement approaches for understanding how model capabilities emerge and evolve during RL training
• Lead strategic evaluation coverage across the company
• Shape the evaluation narrative for model releases
• Lead and mentor a small team of researchers and research engineers, setting research direction and fostering a culture of rigorous, creative research
• Design evaluation frameworks that balance scientific rigor with the practical demands of production training schedules
• Build and maintain relationships across Anthropic's research organization to ensure evaluation insights inform training and deployment decisions
• Contribute to the broader research community through publications, open-source contributions, or external engagement on evaluation best practices

You may be a good fit if you: 

• Have significant experience designing and running evaluations for large language models or similar complex ML systems
• Have led technical projects or teams, either formally or through sustained ownership of critical research directions
• Are equally comfortable designing experiments and writing code—you can move between research and implementation fluidly
• Think strategically about what to measure and why, not just how to measure it
• Can synthesize information across multiple teams and workstreams to form a coherent picture of model capabilities
• Communicate complex technical findings clearly to both technical and non-technical audiences
• Are results-oriented and thrive in fast-paced environments where priorities shift based on research findings
• Care deeply about AI safety and want your work to directly influence how capable AI systems are developed and deployed

Strong candidates may also have: 

• Experience building evaluations for long-horizon or agentic tasks
• Deep familiarity with Reinforcement Learning training dynamics and how model behavior changes during training
• Published research in machine learning evaluation, benchmarking, or related areas
• Experience with safety evaluation frameworks and red teaming methodologies
• Background in psychometrics, experimental psychology, or other measurement-focused disciplines
• A track record of communicating evaluation results to inform high-stakes decisions about model development or deployment
• Experience managing or mentoring researchers and engineers

Representative projects: 

• Designing and implementing a suite of long-horizon evaluations that test model capabilities on tasks requiring sustained reasoning, planning, and tool use over extended interactions
• Building systems to track capability development across RL training checkpoints, surfacing insights about when and how specific capabilities emerge
• Conducting a cross-org audit of evaluation coverage, identifying blind spots, and prioritizing new evaluations to fill critical gaps across Pretraining, RL, Inference, and Product
• Developing the evaluation methodology and narrative for a major model release, working with research leads and communications to clearly characterize model capabilities and limitations
• Researching and prototyping novel evaluation approaches for capabilities that are difficult to measure with existing benchmarks
• Leading a team effort to build reusable evaluation infrastructure that serves multiple teams across the research organization

Why this role matters

AI capabilities are advancing faster than developers can keep up. The gap between what Claude can do and what developers know how to build with it is widening with every model release, every new API feature, and every capability unlock. Closing that gap isn’t just a content problem—it’s a strategic imperative that directly affects how this technology gets adopted, integrated, and ultimately trusted in production systems around the world.

Anthropic’s Ministry of Education exists to help people and organizations develop genuine fluency with AI—not just product training, but the kind of understanding that compounds over time and transfers as capabilities advance. This role owns the developer side of that mission: creating and managing the roadmap for all developer-facing education content and distribution.

The role

As developer education lead, you’ll own the end-to-end strategy, roadmap, and execution for how developers learn to build with Claude. This is not a traditional content role. You’ll sit at the intersection of DevRel, product, and education—deeply embedded in product launches, working cross-functionally with engineering, developer relations, GTM, and the broader education team to ensure every major capability ships with the educational content developers need to adopt it. This is a technical role, with an expectation of using and testing dev tools in your workflows.

You’ll build and own a content pipeline that moves from idea to validation to production to launch—replacing ad-hoc, vibes-based prioritization with a system that’s rigorous about what gets made and why. Your scope spans video content, written tutorials, API courses, cookbooks, and whatever formats emerge as Claude’s capabilities evolve. You’ll work closely with our DevRel team, video producers, and technical writers to create content that helps developers go from “I’ve heard of Claude” to “I’m building production systems with it.”

Key responsibilities

• Own the developer education content roadmap—prioritizing what gets built, when, and for whom across video, written, and course formats

• Partner deeply with product and engineering teams on launches—ensuring every major feature, API update, and capability unlock ships with developer education content + documentation

• Collaborate with DevRel to align on YouTube strategy, community content, and developer community programs (e.g., ambassador programs, public certifications)

• Manage and coordinate with video producers, technical writers, and contractors to execute against the roadmap

• Conduct ongoing content audits—identifying what to update, remove, and add as Claude’s capabilities evolve (e.g., new API features, deprecations, agentic capabilities)

• Define and track content performance metrics that go beyond vanity metrics—measuring actual developer adoption, activation, and fluency gains

• Work with cross-functional teams in MoE, marketing and product to ensure education reaches developer audiences when and where they need it and has sufficient parity with offerings across non-technical audiences

You might be a good fit if

• You have 5+ years in developer content strategy, developer education, DevRel, or technical content leadership—ideally at a developer tools or platform company

• You’ve owned a content roadmap end-to-end and can point to systems you built for prioritizing, producing, and measuring technical content at scale

• You’re deeply cross-functional—comfortable working across engineering, product, marketing, and community teams and translating between their priorities

• You understand developers as an audience, because you are one. You're experienced building with developer tools and can comfortably jump in to use and understand new dev tools independently

• You’re a strong writer and editor who can translate complex technical concepts without dumbing them down

• You use AI tools (ideally Claude) extensively in your own work—not as a novelty, but as core infrastructure for how you think, write, and ship

• You’re comfortable building processes from scratch in fast-moving, ambiguous environments where the product itself is changing rapidly

• You have a strong perspective on how developer education needs to evolve as AI capabilities advance toward AGI

Familiarity with Python is required for this role.

Strong candidates may also have

• Experience managing video content programs (YouTube channels, course platforms, tutorial series)

• Background in API documentation, developer experience, or technical writing

• Familiarity with AI/ML concepts and the developer tooling landscape

• Experience with developer education platforms (Skilljar, Coursera, or similar)

• Prior experience at a high-growth technology company navigating rapid product evolution

About the Role

About the Role

Anthropic's Human Data Interfaces team builds the systems that collect data to improve our models. This includes novel interfaces for data vendors, tooling, and front-end and back-end infrastructure that enables researchers to gather high-quality data at scale. As a Software Engineer, you'll own the architecture and execution of our data collection pipelines — designing systems that are both performant at scale and resilient to the rapidly changing needs of our research teams. You'll work closely with researchers, our cross-functional data operations partners, and the crowdworkers and vendors who use these tools day-to-day.

Responsibilities:

• Architect and build data collection pipelines that support rapid iteration, balancing data quality and system maintainability

• Think deeply about the experience of the crowdworkers and vendors using these systems, building interfaces that are clear, efficient, and lead to high-quality data

• Collaborate closely with research teams to understand evolving data needs and iterate quickly on collection methods

• Partner with our Human Data Operations team to understand the end-to-end data workflow and design interfaces that make their jobs easier

• Prioritize and juggle multiple workstreams, making trade-off decisions in a fast-moving environment where research priorities can shift quickly

You May Be a Good Fit If You:

• Believe that advanced AI systems could have a transformative effect on the world and are interested in helping make sure that transformation goes well

• Are a strong full-stack engineer with broad experience across the stack

• Are very good at building internal tools, including working with users of the tools to understand their needs

• Thrive in fast-moving environments where you need to balance speed of iteration with long-term system health

• Are a quick study—this team sits at the intersection of a large number of different complex technical systems that you'll need to understand (at a high level) to be effective

Strong Candidates May Also Have:

• Experience building human data labelling interfaces, human-in-the-loop systems, or data collection pipelines

• Familiarity with how preference data and reward models are used in AI model training

• Experience working with researchers who are internal users/customers 

• Background in building, and improving the user-experience of user-facing applications, particularly those involving complex UI interactions or annotation workflows

• Strong instincts around system design — building things that evolve gracefully as requirements change

• Experience influencing technical and product direction on a team

About the Team

The Frontier Red Team (FRT) is a technical research team within Anthropic’s Policy organization. Our goal is to make the entire world safer in this era of advanced AI by understanding what these systems can do and building the defenses that matter.

In 2026, we're focused on researching and ensuring safety with self-improving, highly autonomous AI systems—especially ones with cyberphysical capabilities. See our previous related work on cyberdefense, robotics, and Project Vend. We'll also be collaborating closely with the Autonomy workstream to understand novel, societal-scale risks that arise when agents interface with the external world. This is early-stage, high-conviction research with the potential for outsized impact. 

Note: We are exclusively hiring in SF. We support relocation, but all hires must relocate before starting.

About the Role

This Research Scientist will focus on scoping, evaluating, red teaming, and defending against societal risks caused by advanced models that emerge over the next few years. Powerful AI models may have major implications for national security, running a business, power and privacy, infrastructure, social relationships, and more. They may come as a result of the increasing integration of powerful models in our economy and social sphere.

As an independent Research Scientist, you’ll build a research program to understand these Emerging Risks. You’ll build evals, run experiments, and look for real world signals to understand how these may come about. You’ll turn this into insights we can use to steer the development and use of the technology more positively. Compared to the team's other focuses, you will focus less on acute catastrophic risks and more on risks that emerge from increasing integration into our world.

What You’ll Do:

• Design and run research experiments to understand the emerging risks models may create

• Produce internal & external artifacts (research, products, demos, dashboards, tools) that communicate the state of model capabilities

• Shape product,  safeguards, and training decisions based on what you find

• Work closely with Societal Impacts (SI) and Safeguards teams

Sample Projects:

• Build, run, and study an autonomous AI-powered business (e.g. Project Vend), then identify the growth of real autonomous businesses in the wild using Clio and other tools

• Build a benchmark for a model’s national security capabilities

• Red team unsafeguarded models’ abilities to be used for control

• Identify indicators of models being used to scale movements that rely on social control

You May Be a Good Fit If You:

• Are a fast experimentalist who ships research quickly

• Have experience creating a research program from scratch

• Are thoughtful about humanity’s adaptation to powerful AI systems in our economy and society

• Can communicate thoughtfully in written + spoken form with a wide range of stakeholders

• Can scope ambiguous research questions into tractable first projects

Strong candidates may also have experience with:

• Building & maintaining large, foundational infrastructure

• Building simple interfaces that allow non-technical collaborators to evaluate AI systems

• Working with and prioritizing requests from a wide variety of stakeholders, including research and product teams

About the role

We believe skill with AI is fundamental to human agency. Education Labs builds the paradigms that help people become genuinely more capable—not just more engaged.

This is a new kind of role: part researcher, part product builder, part interaction designer. You'll be the second technical builder on a small team studying how AI transforms human capability—and shipping features based on what we discover. You'll have significant creative license to define what "good" looks like, exploring new interaction patterns rather than optimizing existing ones.

We're skeptical of tutorials, onboarding flows, and engagement metrics. We care about experiences that make users progressively more capable, curious, and empowered over time. This means integrating skill development into product design, using Claude itself as a capability-building partner, and measuring success by how users actually grow.

You'll operate as a one-person technical shop: prototyping new ideas, establishing technical direction, and shipping production-quality features to millions of users. You'll need strong product instincts and clean interface design sensibilities, balanced with comfort in ambiguity and frontier thinking.

Responsibilities:

• Ship features that help users develop real skill with AI—measuring success by capability growth, not time-on-site

• Architect end-to-end prototypes (front-end and back-end) that test new interaction paradigms, with particular attention to the front-of-the-frontend: motion, polish, and interaction feel

• Define technical direction for the team—establish patterns others can follow

• Build relationships across Product, Design, and Research to influence how skill development principles shape Anthropic's broader product strategy

• Shape team strategy and roadmap—identify the highest-leverage opportunities and build conviction across stakeholders

• Translate research insights about skill development and human-AI collaboration into shipped product through close collaboration with researchers

• Document and share your work through clear writing, prototypes, and presentations that influence thinking across the organization

You may be a good fit if you have:

Strong full-stack engineering with design sensibility

• 6+ years building and shipping web products, with deep expertise across the stack

• Strong front-end craft: TypeScript/JavaScript, React, CSS—with an eye for interaction design, motion, and visual detail

• Solid back-end and data pipeline experience: Python, API design, analytics infrastructure

• A portfolio showcasing innovative interaction designs and high-quality implementations

• Track record of independently driving features from prototype to production

Deep conviction about human capability

• Strong perspective on how technology should enhance human capabilities rather than diminish them

• Experience or genuine passion for skill development, HCI, developer tools, or products that help people become more capable

• Skepticism of purely engagement-driven metrics; interest in measuring capability outcomes

Research mindset with product execution

• Comfort with ambiguity and exploring undefined problem spaces

• Ability to rapidly prototype, test with users, and iterate toward production

• Strong instincts for product design and user experience, even without formal design training

Strategic leadership and coalition building

• Experience setting vision, shaping team strategy, and building conviction across cross-functional stakeholders

• Ability to build productive relationships with Product, Design, Research, and Engineering teams—especially when your team isn't the owner

• Strong sense of prioritization—knowing what to build now, what to defer, and what to cut

• Track record of influencing roadmaps and decisions beyond your immediate team

Strong candidates may also have:

• Experience in developer tools, creative tools, learning platforms, or other products where user skill development and mastery matter more than time-on-site

• Background in learning sciences, cognitive science, HCI, skill acquisition research, or educational psychology (formal or self-directed)

• Experience with experimentation frameworks, A/B testing, or analytics that measure capability development in production

• Previous experience in research labs, frontier tech companies, or startups with high autonomy and ambiguity

• Published writing, talks, or open-source work on skill development, human-AI interaction, or product philosophy

• Experience building AI-native product experiences or working with LLMs in production contexts

Strong candidates may also have

• Background in learning sciences, cognitive science, HCI, or educational psychology

• Experience in developer tools, creative tools, or learning platforms where mastery matters more than engagement

• Published writing, talks, or open-source work on skill development or human-AI interaction

• Experience building AI-native product experiences or working with LLMs in production

What this role is not

This is a hands-on technical role building product features, embedded within a research team. You'll provide technical guidance and help set direction, but this role doesn't involve people management off the bat. If you're looking to immediately transition into engineering management or lead a large team, this likely isn't the right fit.

About the role

As a Societal Impacts research scientist on the Models Research Pod, you'll close the loop between observing Claude's behavior and improving it at the model level. You'll use observational tools like Clio to analyze real-world usage patterns and build evaluations that  assess whether Claude provides safe responses aligned with its Constitution.

Strong candidates will have experience with machine learning systems and a genuine interest in societal impacts research. You should be adaptable and excited to contribute to evolving team priorities rather than coming in with a fixed agenda. The role is highly cross-functional, with regular collaboration across the fine-tuning, safeguards, policy, and interpretability teams.

We're hiring at both junior and senior levels. Senior researchers should be comfortable doing hands-on technical work alongside helping set research direction.

For this role, we are open to a 6-month residency at Anthropic with the expectation that residents convert to full-time if they do well in their residency. We are exclusively hiring in SF. We support relocation, but all hires must relocate before starting.

Responsibilities:

• Using observational tools like Clio to analyze real-world usage patterns and surface insights about how people interact with Claude.
• Building and running evaluations to assess Claude's behavior across key dimensions of its Constitution, such as safety and quality of advice in high-stakes situations.
• Partnering closely with fine-tuning, safeguards, policy, and interpretability teams to translate research insights into model improvements.
• Generating insights about the societal impact of Anthropic's systems and using this understanding to inform company strategy, research priorities, and policy positions.
• Sharing your work through research publications and external presentations, and developing tools and frameworks that make AI systems more understandable to policymakers, academics, and civil society.

You may be a good fit if:

• You have experience working with machine learning systems and are comfortable with technical infrastructure for interfacing with models.
• You have an interest in societal impacts research; prior experience in this area is a plus but not required.
• You're adaptable and collaborative, able to take direction and contribute to team priorities rather than needing to pursue a predetermined research agenda.
• You're skilled at writing up and communicating your results, even when they're null or unexpected.
• You find it exciting to partner with colleagues across teams on large-scale projects where the whole company contributes to building and analyzing AI systems.
• You have a background in machine learning, data science, or another technical field that involves generating insights from complex systems.
• You're passionate about translating research insights into actionable recommendations for improving AI systems and informing policy.
• For senior candidates: you're willing to do hands-on work while also helping shape research direction, and you may be interested in management opportunities down the line.

Some examples of our work:

• Clio: A system for privacy-preserving insights into real-world AI use

• Values in the Wild

• How People Use Claude for Support, Advice, and Companionship

• Economic Index

• How AI is Transforming Work at Anthropic

• Education Report: How University Students Use Claude

• The Capacity for Moral Self-Correction in Large Language Models

• Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned

• Challenges in evaluating AI systems

• Towards Measuring the Representation of Subjective Global Opinions in Language Models

• Collective Constitutional AI: Aligning a Language Model with Public Input

About the Team

The Frontier Red Team (FRT) is a small, focused technical research team within Anthropic's Policy organization. Our goal is to make the entire world safer in this era of advanced AI by understanding what these systems can do and building the defenses that matter.

In 2026, we're focused on researching and ensuring safety with self-improving, highly autonomous AI systems—especially ones with cyberphysical capabilities. See our previous related work on cyberdefense, robotics, and Project Vend. We'll also be collaborating closely with the Emerging Risks workstream to understand novel, societal-scale risks that arise when agents interface with the external world. This is early-stage, high-conviction research with the potential for outsized impact. 

Note: We are exclusively hiring in SF. We support relocation, but all hires must relocate before starting.

About the Role

Our team is focused on a critical question: how do we defend against a world where powerful, autonomous, self-improving AI systems may be used adversarially?

As a Research Engineer on our team, you'll build and eval model organisms of autonomous systems and develop the defensive agents needed to counter them. This work sits at the intersection of AI capabilities research, security, and policy—what we learn directly shapes how Anthropic and the world prepare for advanced AI.

This is applied research with real-world stakes. Your work will inform decisions at the highest levels of the company, contribute to public demonstrations that shape policy discourse, and help build technical defenses that could matter enormously as AI systems become more capable.

What You'll Do

• Design and build autonomous AI systems that can use tools and operate across diverse environments—creating model organisms that help us understand and defend against advanced adversarial AI
• Create evals and training environments to understand and shape agent behavior in desirable ways
• Develop defensive agents that can detect, disrupt, or outcompete adversarial AI systems in realistic scenarios
• Interface Claude with hardware platforms (e.g. robotics, physical systems) to understand cyberphysical risks and defenses
• Translate technical findings into compelling demonstrations and artifacts that inform policymakers and the public
• Collaborate with external experts in cybersecurity, national security, and AI safety to scope and validate research directions

Sample Projects

• Developing systems where Claude controls diverse hardware and robotics platforms simultaneously
• Creating attack-defend simulations (CTFs, wargames, adversarial games) to test defensive AI capabilities
• Designing and implementing RL environments for training defensive agents
• Pointing autonomous systems at real-world security challenges to characterize risks and develop mitigations

You May Be a Good Fit If You

• Have strong software engineering skills, particularly in Python
• Have experience building and working with LLM-based agents or autonomous systems
• Are driven to find solutions to ambiguously scoped, high-stakes problems
• Design and run experiments quickly, iterating fast toward useful results
• Thrive in collaborative environments (we love pair programming!)
• Care deeply about AI safety and want your work to have real-world impact on how humanity navigates advanced AI
• Can own entire problems end-to-end, including both technical and non-technical components
• Are comfortable working on sensitive projects that require discretion and integrity

Strong Candidates May Also Have

• Experience with reinforcement learning, self-play, or multi-agent systems
• Experience with robotics, hardware interfaces, or cyberphysical systems
• Track record of building demos or prototypes that communicate complex technical ideas
• Experience working with external stakeholders (policymakers, government, researchers)
• Familiarity with AI safety research and threat modeling for advanced AI systems

About the Team

The Universes team within Research is responsible for training AI models to perform complex, difficult, long-horizon agentic tasks in ultra-realistic settings. We design and implement novel training environments that go far beyond what models can do today — environments where models learn to navigate ambiguity, handle interruptions, maintain context over extended interactions, and exercise judgment in open-ended scenarios.

About the Role

We're looking for Research Engineers to help us build the next generation of training environments for capable and safe agentic AI.

This role blends research and engineering responsibilities, requiring you to both implement novel approaches and contribute to research direction. You'll work on fundamental research in reinforcement learning, designing training environments and methodologies that push the state of the art, and building evaluations that measure genuine capability.

Responsibilities:

• Build the next generation of agentic environments

• Build rigorous evaluations that measure real capability

• Collaborate across research and infrastructure teams to ship environments into production training

• Debug and iterate rapidly across research and production ML stacks

• Contribute to research culture through technical discussions and collaborative problem-solving

You may be a good fit if you:

• Are highly impact-driven — you care about outcomes, not activity

• Operate with high agency

• Have good research taste or senior technical experience, demonstrating good judgment in identifying what actually matters in complex problem spaces

• Can balance research exploration with engineering implementation

• Are passionate about the potential impact of AI and are committed to developing safe and beneficial systems

• Are comfortable with uncertainty and adapt quickly as the landscape shifts

• Have strong software engineering skills and can build robust infrastructure

• Enjoy pair programming (we love to pair!)

Strong candidates may also have one or more of the following:

• Have industry experience with large language model training, fine-tuning or evaluation

• Have industry experience building RL environments, simulation systems, or large-scale ML infrastructure

• Senior experience in a relevant technical field even if transitioning domains

• Deep expertise in sandboxing, containerization, VM infrastructure, or distributed systems

• Published influential work in relevant ML areas

About Horizons

The Horizons team leads Anthropic's reinforcement learning (RL) research and development, playing a critical role in advancing our AI systems. We've contributed to every Claude release, with significant impact on the autonomy, coding, and reasoning capabilities of Anthropic's models.

About the role

We're hiring for the Cybersecurity RL team within Horizons. As a Research Engineer, you'll help to safely advance the capabilities of our models in secure coding, vulnerability remediation, and other areas of defensive cybersecurity.

This role blends research and engineering, requiring you to both develop novel approaches and realize them in code. Your work will include designing and implementing RL environments, conducting experiments and evaluations, delivering your work into production training runs, and collaborating with other researchers, engineers, and cybersecurity specialists across and outside Anthropic.

The role requires domain expertise in cybersecurity paired with interest or experience in training safe AI models. For example, you might be a white hat hacker who's curious about how LLMs could augment or transform your work, a security engineer interested in how AI could help harden systems at scale, or a detection and response professional wondering how models could enhance defensive workflows.

You may be a good fit if you:

• Have experience in cybersecurity research.

• Have experience with machine learning.

• Have strong software engineering skills.

• Can balance research exploration with engineering implementation.

• Are passionate about AI's potential and committed to developing safe and beneficial systems.

Strong candidates may also have:

• Professional experience in security engineering, fuzzing, detection and response, or other applied defensive work.

• Experience participating in or building CTF competitions and cyber ranges.

• Academic research experience in cybersecurity.

• Familiarity with RL techniques and environments.

• Familiarity with LLM training methodologies.

About the role

As a Senior Research Scientist on our Reward Models team, you'll lead research efforts to improve how we specify and learn human preferences at scale. Your work will directly shape how our models understand and optimize for what humans actually want — enabling Claude to be more useful, more reliable, and better aligned with human values.

This role focuses on pushing the frontier of reward modeling for large language models. You'll develop novel architectures and training methodologies for RLHF, research new approaches to LLM-based evaluation and grading (including rubric-based methods), and investigate techniques to identify and mitigate reward hacking. You'll collaborate closely with teams across Anthropic, including Finetuning, Alignment Science, and our broader research organization, to ensure your work translates into concrete improvements in both model capabilities and safety. 

We're looking for someone who can drive ambitious research agendas while also shipping practical improvements to production systems. You'll have the opportunity to work on some of the most important open problems in AI alignment, with access to frontier models and significant computational resources. Your work will directly advance the science of how we train AI systems to be both highly capable and safe. 

Note: For this role, we conduct all interviews in Python.

Responsibilities

• Lead research on novel reward model architectures and training approaches for RLHF
• Develop and evaluate LLM-based grading and evaluation methods, including rubric-driven approaches that improve consistency and interpretability
• Research techniques to detect, characterize, and mitigate reward hacking and specification gaming
• Design experiments to understand reward model generalization, robustness, and failure modes
• Collaborate with the Finetuning team to translate research insights into improvements for production training pipelines
• Contribute to research publications, blog posts, and internal documentation
• Mentor other researchers and help build institutional knowledge around reward modeling

You may be a good fit if you

• Have a track record of research contributions in reward modeling, RLHF, or closely related areas of machine learning
• Have experience training and evaluating reward models for large language models
• Are comfortable designing and running large-scale experiments with significant computational resources
• Can work effectively across research and engineering, iterating quickly while maintaining scientific rigor
• Enjoy collaborative research and can communicate complex ideas clearly to diverse audiences
• Care deeply about building AI systems that are both highly capable and safe

Strong candidates may also

• Have published research on reward modeling, preference learning, or RLHF
• Have experience with LLM-as-judge approaches, including calibration and reliability challenges
• Have worked on reward hacking, specification gaming, or related robustness problems
• Have experience with constitutional AI, debate, or other scalable oversight approaches
• Have contributed to production ML systems at scale
• Have familiarity with interpretability techniques as applied to understanding reward model behavior

About the role

You will deeply understand the research workflows of our Finetuning teams and automate the high-friction parts – turning days of manual experimentation into hours. You’ll build the tools and infrastructure that enable researchers across the organization to develop, evaluate, and optimize reward signals for training our models. Your scalable platforms will make it easy to experiment with different reward methodologies, assess their robustness, and iterate rapidly on improvements to help the rest of Anthropic train our reward models.

This is a role for someone who wants to stay close to the science while having outsized leverage. You'll partner directly with researchers on the Rewards team and across the broader Fine-Tuning organization to understand what slows them down: running human data experiments before adding to preference models, debugging reward hacks, comparing rubric methodologies across domains. Then you'll build the systems that make those workflows 10x faster. When you have bandwidth, you'll contribute directly to research projects yourself. Your work will directly impact our ability to scale reward development across domains, from crafting and evaluating rubrics to understanding the effects of human feedback data to detecting and mitigating reward hacks. 

We're looking for someone who combines strong engineering fundamentals with research experience – someone who can scope ambiguous problems, ship quickly, and cares as much about the science as the systems.

Note: For this role, we conduct all interviews in Python.

Responsibilities

• Design and build infrastructure that enables researchers to rapidly iterate on reward signals, including tools for rubric development, human feedback data analysis, and reward robustness evaluation
• Develop systems for automated quality assessment of rewards, including detection of reward hacks and other pathologies
• Create tooling that allows researchers to easily compare different reward methodologies (preference models, rubrics, programmatic rewards) and understand their effects
• Build pipelines and workflows that reduce toil in reward development, from dataset preparation to evaluation to deployment
• Implement monitoring and observability systems to track reward signal quality and surface issues during training runs
• Collaborate with researchers to translate science requirements into platform capabilities
• Optimize existing systems for performance, reliability, and ease of use
• Contribute to the development of best practices and documentation for reward development workflows

You may be a good fit if you

• Have prior research experience 
• Are excited to work closely with researchers and translate ambiguous requirements into well-scoped engineering projects
• Have strong Python skills
• Have experience with ML workflows and data pipelines, and building related infrastructure/tooling/platforms
• Are comfortable working across the stack, ranging from data pipelines to experiment tracking to user-facing tooling
• Can balance building robust, maintainable systems with the need to move quickly in a research environment
• Are results-oriented, with a bias towards flexibility and impact
• Pick up slack, even if it goes outside your job description
• Care about the societal impacts of your work and are motivated by Anthropic's mission to develop safe AI

Strong candidates may also have experience with

• Experience with ML research
• Building internal tooling and platforms for ML researchers
• Data quality assessment and pipeline optimization
• Experiment tracking, evaluation frameworks, or MLOps tooling
• Large-scale data processing (e.g., Spark, Hive, or similar)
• Kubernetes, distributed systems, or cloud infrastructure
• Familiarity with reinforcement learning or fine-tuning workflows 

Representative projects

• Building infrastructure that allows researchers to rapidly test new rubric designs against small models before scaling up
• Developing automated systems to detect reward hacks and surface problematic behaviors during training
• Creating tooling for comparing different grading methodologies and understanding their effects on model behavior
• Building a data quality flywheel that helps researchers identify problematic transcripts and feed improvements back into the system
• Developing dashboards and monitoring systems that give researchers visibility into reward signal quality across training runs
• Streamlining dataset preparation workflows to reduce latency and operational overhead

About the role:

When you see what modern language models are capable of, do you wonder, "How do these things work? How can we trust them?"

The Interpretability team at Anthropic is working to reverse-engineer how trained models work because we believe that a mechanistic understanding is the most robust way to make advanced systems safe.

Think of us as doing "neuroscience" of neural networks using "microscopes" we build - or reverse-engineering neural networks like binary programs.

More resources to learn about our work: 

• Our research blog - covering advances including Monosemantic Features and Circuits
• An Introduction to Interpretability from our research lead, Chris Olah
• The Urgency of Interpretability from CEO Dario Amodei
• Engineering Challenges Scaling Interpretability - directly relevant to this role
• 60 Minutes segment - Around 8:07, see a demo of tooling our team built
• New Yorker article - what it's like to work on one of AI's hardest open problems

Even if you haven’t worked on interpretability before, the infrastructure expertise is similar to what's needed across the lifecycle of a production language model:

• Pretraining: Training dictionary learning models looks a lot like model pretraining - creating stable, performant training jobs for massively parameterized models across thousands of chips
• Inference: Interp runs a customized inference stack. Day-to-day analysis requires services that allow editing a model's internal activations mid-forward-pass - for example, adding a "steering vector"
• Performance: Like all LLM work, we push up against the limits of hardware and software. Rather than squeezing the last 0.1%, we are focused on finding bottlenecks, fixing them and moving ahead given rapidly evolving research and safety mission

The science keeps scaling - and it's now applied directly in safety audits on frontier models, with real deadlines. As our research has matured, engineering and infrastructure have become a bottleneck. Your work will have a direct impact on one of the most important open problems in AI.

Responsibilities:

• Build and maintain the specialized inference and training infrastructure that powers interpretability research - including instrumented forward/backward passes, activation extraction, and steering vector application
• Resolve scaling and efficiency bottlenecks through profiling, optimization, and close collaboration with peer infrastructure teams
• Design tools, abstractions, and platforms that enable researchers to rapidly experiment without hitting engineering barriers
• Help bring interpretability research into production safety audits - with real deadlines and high reliability expectations
• Work across the stack - from model internals and accelerator-level optimization to user-facing research tooling

You may be a good fit if you:

• Have 5-10+ years of experience building software
• Are highly proficient in at least one programming language (e.g., Python, Rust, Go, Java) and productive with Python
• Are extremely curious about unfamiliar domains; can quickly learn and put that knowledge to work, e.g. diving into new layers of the stack to find bottlenecks
• Have a strong ability to prioritize the most impactful work and are comfortable operating with ambiguity and questioning assumptions
• Prefer fast-moving collaborative projects to extensive solo efforts
• Are curious about interpretability research and its role in AI safety (though no research experience is required!)
• Care about the societal impacts and ethics of your work
• Are comfortable working closely with researchers, translating research needs into engineering solutions.

Strong candidates may also have experience with:

• Optimizing the performance of large-scale distributed systems
• Language modeling fundamentals with transformers
• High Performance LLM optimization: memory management, compute efficiency, parallelism strategies, inference throughput optimization
• Working hands-on in a mainstream ML stack - PyTorch/CUDA on GPUs or JAX/XLA on TPUs
• Collaborating closely with researchers and building tooling to support research teams; or directly performed research with complex engineering challenges

Representative Projects:

• Building Garcon, a tool that allows researchers to easily instrument LLMs to extract internal activations
• Designing and optimizing a pipeline to efficiently collect petabytes of transformer activations and shuffle them
• Profiling and optimizing ML training jobs, including multi-GPU parallelism and memory optimization
• Building a steered inference system that applies targeted interventions to model internals at scale (conceptually similar to Golden Gate Claude but for safety research)

Role Specific Location Policy:

• This role is based in the San Francisco office; however, we are open to considering exceptional candidates for remote work on a case-by-case basis.

Note: we don't have open Research Manager positions on the Interpretability team at this time. However, we're actively growing our team of Research Engineers and Research Scientists. If you're excited about interpretability research and open to an individual contributor role, we encourage you to apply.

About the Interpretability team:

When you see what modern language models are capable of, do you wonder, "How do these things work? How can we trust them?"

The Interpretability team’s mission is to reverse engineer how trained models work, and Interpretability research is one of Anthropic’s core research bets on AI safety. We believe that a mechanistic understanding is the most robust way to make advanced systems safe. 

People mean many different things by "interpretability". We're focused on mechanistic interpretability, which aims to discover how neural network parameters map to meaningful algorithms. Some useful analogies might be to think of us as trying to do "biology" or "neuroscience" of neural networks, or as treating neural networks as binary computer programs we're trying to "reverse engineer".

We aim to create a solid scientific foundation for mechanistically understanding neural networks and making them safe (see our vision post). We have focused on resolving the issue of "superposition" (see Toy Models of Superposition, Superposition, Memorization, and Double Descent, and our May 2023 update), which causes the computational units of the models, like neurons and attention heads, to be individually uninterpretable, and on finding ways to decompose models into more interpretable components. Our subsequent work which found millions of features in Claude 3.0 Sonnet, one of our production language models, represents progress in this direction. In our most recent work, we developed methods that allow us to build circuits using features and use these circuits to understand the mechanisms associated with a model's computation and study specific examples of multi-hop reasoning, planning, and chain-of-thought faithfulness on Claude Haiku 3.5, one of our production models.” This is a stepping stone towards our overall goal of mechanistically understanding neural networks.

A few places to learn more about our work and team are this introduction to Interpretability from our research lead, Chris Olah, Stanford CS25 lecture given by Josh Batson, and TWIML AI podcast with Emmanuel Ameisen.

Some of our team's notable publications include and our Circuits’ Methods and Biology papers, Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet, Towards Monosemanticity: Decomposing Language Models With Dictionary Learning, A Mathematical Framework for Transformer Circuits, In-context Learning and Induction Heads, and Toy Models of Superposition. This work builds on ideas from members' work prior to Anthropic such as the original circuits thread, Multimodal Neurons, Activation Atlases, and Building Blocks.

About the role:

As a manager on the Interpretability team, you'll support a team of expert researchers and engineers who are trying to understand at a deep, mechanistic level, how modern large language models work internally. 

Few things can accelerate this work more than great managers. Your work as manager will be critical in making sure that our fast-growing team is able to meet its ambitious safety research goals over the coming years. In this role, you will partner closely with an individual contributor research lead to drive the team's success, translating cutting-edge research ideas into tangible goals and overseeing their execution. You will manage team execution, careers and performance, facilitate relationships within and across teams, and drive the hiring pipeline. 

If you're more interested in making individual direct technical contributions to our research as the primary focus of your role, feel free to apply to our Research Scientist or Research Engineer roles instead.

Responsibilities:

• Partner with a research lead on direction, project planning and execution, hiring, and people development
• Set and maintain a high bar for execution speed and quality, including identifying improvements to processes that help the team operate effectively 
• Coach and support team members to have more impact and develop in their careers
• Drive the team's recruiting efforts, including hiring planning, process improvements, and sourcing and closing
• Help identify and support opportunities for collaboration with other teams across Anthropic
• Communicate team updates and results to other teams and leadership
• Maintain a deep understanding of the team's technical work and its implications for AI safety

You may be a good fit if you:

• Are an experienced manager (minimum 2-5 years) with a track record of effectively leading highly technical research and/or engineering teams 
• Have a background in machine learning, AI, or a related technical field
• Actively enjoy people management and are experienced with coaching and mentorship, performance evaluation, career development, and hiring for technical roles
• Have strong project management skills, including prioritization and cross-functional coordination and collaboration
• Have managed technical teams through periods of ambiguity and change
• Are a quick learner, capable of understanding and contributing to discussions on complex technical topics and are motivated to learn about our research
• Are a strong communicator both in speaking and in writing
• Believe that advanced AI systems could have a transformative effect on the world, and are passionate about helping make sure that transformation goes well

Strong candidates may also have:

• Experience scaling engineering infrastructure
• Experience working on open-ended, exploratory research agendas aimed at foundational insights
• Some familiarity with our work and mechanistic interpretability

Role Specific Location Policy:

• This role is expected to be in our SF office for 3 days a week.