Frontier Risk Report (February to March 2026) - METR
A pilot assessment of rogue deployment risk at frontier AI companies. Starting in February 2026, METR conducted a pilot exercise to assess misalignment risks from AI agents used inside frontier AI developers, with participation from Anthropic, Google, Meta, and OpenAI.
Frontier Risk Report (February to March 2026) - METR Our Work Research Notes Updates Risk Assessment About Donate Careers Language English Español 中文 Search --> Our Work Research Notes Updates Risk Assessment About Donate Careers English Español 中文 English Español 中文 Menu × Frontier Risk Report (February to March 2026) DATE May 19, 2026 SHARE Copy Link Citation BibTeX Citation × @misc { metr-2026-frontier-risk-report , title = {Frontier Risk Report (February to March 2026)} , author = {METR} , howpublished = {\url{https://metr.org/blog/2026-05-19-frontier-risk-report/}} , year = {2
Explore this link on the map →saved by
related reading
- METRmetr.org
- AI 2027ai-2027.com
- Current AIs seem pretty misaligned to me — LessWronglesswrong.com
- About METRmetr.org
- AI 2027ai-2027.com
- AI in 2025: gestalt — LessWronglesswrong.com
- Current AIs seem pretty misaligned to meblog.redwoodresearch.org
- Measuring AI Ability to Complete Long Tasks - METRmetr.org
- The Rogue Replication Threat Model - METRmetr.org
- We spent 2 hours working in the future - METRmetr.org
- Emerging processes for frontier AI safety - GOV.UKgov.uk
- Off Target | CNAScnas.org