AI Security & Control Researcher

The role

This IC role focuses on threat modeling for coding agents, maintaining a failure mode library, and designing attack scenarios. The person will also adjudicate flagged agent trajectories and red-team the Watcher product.

What you will do

• Develop threat models for coding agents assumed to be misaligned or compromised
• Treat coding agents as insider risks and map kill chains and attack progressions adapted for agentic AI
• Build and maintain coding agent security levels that are concrete, testable and usable internally and externally
• Maintain the library of coding agent failure modes and keep it comprehensive and current
• Stay up to date with academic and industry literature on coding agent and AI security
• Design realistic attack trajectories and exploits for monitor development and backtesting
• Serve as security adjudicator for flagged trajectories, judging genuine attacks versus benign unusual workflows
• Red-team and improve Watcher's monitors and policies and document findings

Required qualifications

• 5+ years in security roles in a hands-on technical capacity, not purely GRC or compliance
• Ability to read code, understand infrastructure and evaluate technical controls
• Direct experience with security research, threat modeling or offensive security
• Comfortable working in a small, fast-moving team and owning ambiguous, open-ended research questions
• Strong written communication for threat models, security levels and failure mode documentation

Nice to have

• Experience with AI/ML systems security, LLM security or AI control research
• Detection engineering, SOC or incident analysis experience
• Familiarity with insider threat programs or insider risk frameworks
• Red teaming or offensive security background
• Formal AI safety research background

Benefits

• Market competitive salary and equity
• Competitive benefits
• Flexible work hours and schedule
• Unlimited vacation
• Unlimited sick leave
• Up to 6 months of paid parental leave
• Comprehensive health, dental and vision insurance
• Retirement savings with competitive employer matching (e.g. 401(k) for US employees)
• Lunch, dinner and snacks provided on workdays
• Paid work trips, including staff retreats, business trips and relevant conferences
• Yearly $1,000 (USD) professional development budget
• Relocation support and visa fees (if applicable)

Before you apply

• Visa sponsorship is offered in both the UK and US, but is not guaranteed for every role or candidate
• The interview process includes a screening interview, a 3-hour take-home test, 3 technical interviews and a final interview with the CEO
• AI tools may be used on the take-home test, and the candidate is responsible for the quality of submitted work
• No leetcode-style general coding interviews

About the company

Apollo Research works with frontier AI companies such as OpenAI, Anthropic, Google and Meta to test models and research scheming. Its coding agent security product, Watcher, is deployed in production.