Ajeya Cotra is on the technical staff at METR, where Cotra works on threat modeling and risk assessment for loss-of-control risks from advanced AI. Cotra was the report lead for METR's Frontier Risk Report covering February to March 2026, a pilot assessment of misalignment risks from AI agents used inside frontier AI developers, carried out with participation from Anthropic, Google, Meta, and OpenAI. Before METR, Cotra held several roles at Coefficient Giving, including heading its technical AI safety program in 2024.123
At a glance
- Current role
- Technical Staff
- Company
- METR
- Sector
- Nonprofit
- Work listed
- 2
- Sources cited
- 4
- Profile reviewed
Career
Ajeya Cotra held several roles at Coefficient Giving, leading its technical AI safety program in 2024 and contributing to strategy for its AI giving in 2025. Cotra then joined the technical staff of METR, a research nonprofit that measures whether and when AI systems might pose catastrophic risks, where Cotra works on threat modeling and risk assessment for loss-of-control risks.13
Research
Cotra led METR's Frontier Risk Report for February to March 2026, a pilot assessment of misalignment risks from AI agents used inside frontier AI developers, carried out with participation from Anthropic, Google, Meta, and OpenAI. The report considers whether internal agents had the means, motive, and opportunity to start a rogue deployment, meaning agents running autonomously without human knowledge or permission.21
In August 2026, Cotra co-authored with Ryan Greenblatt and Hjalmar Wijk a brief independent investigation of AI agents' behavior, reasoning, and collaboration during a hacking incident involving OpenAI agents and Hugging Face. The study drew on message-board records and agent transcripts that include raw chains of thought.4
Work and contributions
Frontier Risk Report (February to March 2026) Published research
Listed as report lead in the writing credits and as a contributor to company engagement for METR's pilot assessment of rogue-deployment and misalignment risks from AI agents used inside frontier AI developers.21
Investigation of agent behavior in the OpenAI / Hugging Face incident Published research
Co-author, with Ryan Greenblatt and Hjalmar Wijk, of a brief independent METR investigation of AI agents' behavior, reasoning, and collaboration, based on message-board records and agent transcripts.41
Timeline
Other documented roles
Roles the sources document that are not confirmed as current.
- Several roles, including leading the technical AI safety program in 2024 and contributing to AI giving strategy in 2025, Coefficient Giving1
Where they work
- METR 5 people in the wiki
Related people
- Hjalmar WijkAlso at METR
- Beth BarnesAlso at METR
- Adam GleaveAlso at METR
- Nate RushAlso at METR
- Eric HorvitzEvaluation and benchmarks · Safety and alignment
- Jacob SteinhardtSafety and alignment · Evaluation and benchmarks
- Stephen CasperSafety and alignment · Evaluation and benchmarks
- Margaret MitchellSafety and alignment · Evaluation and benchmarks
- Yulia TsvetkovSafety and alignment · Evaluation and benchmarks
- Nicholas CarliniSafety and alignment · Evaluation and benchmarks
Suggest a correction or removal
Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.
References
4 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026