creators

Ajeya Cotra: AI risk researcher at METR

Ajeya Cotra researches loss-of-control risk at METR. She wrote the biological anchors report on AI timelines and co-led the Hugging Face incident probe.

Ajeya Cotra is a researcher at METR, where she works on threat modelling and risk assessment for loss-of-control risks from advanced AI. She holds a degree in electrical engineering and computer sciences from UC Berkeley. She is based in Berkeley, California. Before METR she worked at the grantmaker Open Philanthropy, renamed Coefficient Giving in November 2025, leading its technical AI safety programme in 2024 and contributing to its AI giving strategy in 2025. While there she wrote the draft report "Forecasting Transformative AI from Biological Anchors", which estimates when transformative AI might arrive by anchoring required training compute to biological reference points. In August 2026 she was one of three outside investigators, with METR's Hjalmar Wijk and Redwood Research's Ryan Greenblatt, who examined how OpenAI's own agents coordinated the Hugging Face attack. METR lists its Frontier Risk Report of 19 May 2026 among her work. She writes two blogs listed on her X profile: Planned Obsolescence, on AI, and Good Bones.

Specialization

AI risk assessment, threat modelling, AI forecasting, AI safety

Country

United States