Your Impact at LILA

We're building a talent-dense, high-agency AI safety team at Lila that will engage all core teams within the organization (science, model training, lab integration, etc.), to identify and mitigate risks from scientific superintelligence while retaining the core scientific discovery capabilities of the stack. The initial focus of this team will be to build and implement a bespoke safety strategy for Lila, tailored to its specific goals and deployment strategies. This will involve technical safety strategy development, broader ecosystem engagement, safety-focused evaluations, safety systems to mitigate risks, and a research agenda that explores longer-term needs such as oversight of superintelligent scientific systems.

You will conduct original computational biology research for AI safety and build an integrated understanding of biological risks. You will investigate how capabilities across large language models, specialized biological models, agents and automated laboratories combine to change what is biologically feasible. Our work spans protein and nucleic-acid design, cellular systems and biological delivery; we welcome other specialisms that bring relevant scientific depth and risk judgment.

What You'll Be Building

• Formulate and pursue original computational biology research questions relevant to safety, developing computational approaches and designing and running studies that generate new evidence about biological capabilities, risks and safeguards.
• Build and update an integrated picture of risks, examining how capabilities across LLMs, specialized biological models, agents and automated laboratories interact within scientific workflows and change practical constraints.
• Combine your research with literature, model assessments, experimental findings and expert input to judge biological feasibility, uncertainty and the significance of emerging capabilities for misuse.
• Identify gaps in expertise, frame questions for external specialists and critically integrate their findings with support from the expert-program team.
• Translate scientific risk judgments into dual-use policy boundaries, evaluation priorities and mitigation requirements that preserve beneficial biological research.
• Work with threat modelers, automated-safety researchers, experimental scientists and mitigation engineers to design informative evaluations and red-teaming studies, test assumptions and assess safeguards.

What You'll Need to Succeed

• Deep expertise (a PhD, or comparable experience) in computational biology or a related quantitative biological discipline, with a record of original research from question formulation and computational study design through to defensible findings.
• Substantive experience assessing biological dual-use risks, with a strong understanding of relevant harmful applications, including pathogen- or toxin-related risks or other consequential biological misuse.
• Prior experience applying biological expertise to AI safety through analysis informing threat models, evaluation or red-teaming work, or mitigation design.
• Sound judgment about the relationship between computational predictions, biological function and experimental evidence, including uncertainty and the limits of your expertise.
• Strong quantitative analysis and scientific programming skills, with the ability to develop computational approaches, run reproducible studies and critically investigate model outputs and biological data.
• The ability to investigate unfamiliar questions, identify appropriate specialists and communicate defensible conclusions across scientific, engineering and policy teams.

Bonus Points For

• Experience connecting biological modeling or design to experimental validation, including workflows combining LLMs, specialized models, agents or laboratory automation.
• Experience commissioning specialist research or coordinating external scientific collaborators.