*This is a hybrid role for San Jose, CA; Austin, TX; Boston, MA; or Dublin, Ireland

Position Overview

Vectra AI’s Security Research team studies how attackers operate and turns that knowledge into capabilities that help customers find and respond to threats. We are looking for a Senior Security Researcher to help define and investigate the security challenges created by enterprise AI adoption and AI-enabled attackers.

You will study how agents, copilots, AI-connected applications, and their identities can be misused or compromised. You will also examine how attackers use AI to accelerate familiar attack techniques. Your research will connect activity across identity, cloud, and network environments, with a focus on what customers can observe, investigate, and act on. This is an emerging area: you will help test assumptions, identify gaps in visibility, and shape the work that follows.

What You Will Do

• Develop threat models for enterprise AI systems and agent-mediated activity, including delegated access, service and workload identities, cloud agent infrastructure, connectors, and downstream actions.
• Investigate how AI changes attacker operations, including the speed, scale, sequencing, and coordination of reconnaissance, credential abuse, lateral movement, and data theft.
• Recreate realistic attack and misuse scenarios in controlled environments. Collect evidence, map the observable behavior across identity, cloud, and network, and distinguish confirmed signals from hypotheses.
• Assess existing detection coverage, identify telemetry and attribution gaps, and work with Data Science and Engineering to develop and validate new behavioral detections.
• Help connect an agent’s actions to the human, application, identity, workload, tool, and resource involved, even when a single log source shows only part of the chain.
• Translate research into clear findings, attack samples, coverage assessments, and recommendations that Product, Engineering, customer-facing teams, and customers can use.
• Stay close to emerging techniques and real incidents, contribute original research, and help Vectra decide which problems warrant further investment.

What Will You Need 

• Approximately 5 or more years in security research, detection engineering, incident response, threat hunting, red teaming, or a related field, with evidence of independently led technical investigations.
• Strong understanding of attacker behavior and enterprise identity, including authentication, permissions, tokens, service principals, or workload identities.
• Hands-on experience investigating security activity in at least one major cloud platform and the ability to reason across cloud, identity, and network evidence.
• Ability to build repeatable labs or simulations, analyze logs and telemetry, and turn findings into detection requirements or coverage tests.
• Programming or scripting ability, such as Python, for research, data analysis, or test automation.
• Clear technical writing and the judgment to explain what the evidence shows, what remains uncertain, and why it matters to customers.
• Ability to work closely with researchers, data scientists, engineers, and product teams in a fast-changing problem space.

What Will Help You

• Research or operational experience with AI agents, copilots, model services, agent frameworks, tool integrations, or AI-related incidents.
• Familiarity with attacks involving delegated authority, OAuth applications, cloud control planes, prompt injection consequences, or compromised agent infrastructure.
• Experience correlating activity across multiple data sources or measuring detection efficacy against realistic attack sequences.
• Published research, open-source tools, conference talks, or other contributions to the security community.