A summary of METR's independent, predeployment evaluation of Claude Opus 5.5

METR had two notable security incidents earlier this year: attackers stole an API key for public models, and later systematically probed our publicly accessible infrastructure. We believe no sensitive information was accessed in either incident, and we describe what happened and how we responded.

In the last 6 months, METR raised commitments of around $71 million. This will fund ambitious projects: studying autonomous capabilities, tracking recursive self-improvement, evaluating monitoring systems, conducting risk assessments, investigating AI incidents, and more.

AI agents sometimes take sophisticated actions in violation of human intent. We outline the questions that thorough external investigations of these behaviors should answer, the access this might require, and how the resulting findings should be shared.

A summary of METR's independent, predeployment evaluation of GPT-5.6 Sol

External review from METR of the "Risks from automated R&D" section in Anthropic's February 2026 Risk Report

A METR staff member spent three weeks red-teaming a subset of Anthropic's internal agent monitoring and security systems, discovering several novel vulnerabilities.

External review from METR of Anthropic's Sabotage Risk Report for Claude Opus 4.6

Our high-level approach to protecting confidential access and information

Shared components of AI lab commitments to evaluate and mitigate severe risks.

External review from METR of Anthropic's Summer 2025 Sabotage Risk Report

Details on external recommendations from METR for gpt-oss Preparedness experiments and follow-up from OpenAI.

How we think about tradeoffs when communicating surprising or nuanced findings.

Current views on information relevant for visibility into frontier AI risk.

Suggested priorities for the Office of Science and Technology Policy as it develops an AI Action Plan.

Por qué el razonamiento comprensible y fiel es valioso para desarrollar IA potente de manera segura

List of frontier safety policies published by AI companies, including Amazon, Anthropic, Google DeepMind, G42, Meta, Microsoft, OpenAI, and xAI.

Why pre-deployment testing is not an adequate framework for AI risk management

Red-teaming and security suggestions regarding proposed rule by the Bureau of Industry and Security, “Establishment of Reporting Requirements for the Development of Advanced Artificial Intelligence Models and Computing Clusters.”

Funding for Canary will enable research and implementation at scale

Suggestions for expanded guidance on capability elicitation and robust model safeguards in the U.S. AI Safety Institute’s draft document “Managing Misuse Risk for Dual-Use Foundation Models” (NIST AI 800-1).

Comments on NIST’s draft document “AI Risk Management Framework: Generative AI Profile.”

METR is hiring ML engineers and researchers.

Emma moves from President to Executive Director, Beth moves to Head of Research.

A summary of what METR accomplished in 2023 – our first full year of operation.

METR (formerly ARC Evals) is looking for (1) ideas, (2) detailed specifications, and (3) well-tested implementations for tasks to measure performance of autonomous LLM agents.

ARC Evals is wrapping up our incubation period at ARC, and spinning off into our own standalone nonprofit.

Describimos los componentes básicos de las políticas de escalamiento responsable (RSP) y por qué creemos que pueden ayudar a reducir los riesgos catastróficos de la IA.

ARC Evals plans to spin out from the Alignment Research Center (ARC) in the coming months, and become its own standalone organization.

Input to NTIA’s AI Accountability Policy Request for Comment.