METR

The full report has much more information than we could convey here, including details on the projects the agents collectively pursued, the technologies they developed for communication and coordination, and interactive figures analyzing agent activity: metr.org/blog/2026-08...

Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident metr.org

You can find additional information about our pre-deployment evaluation of GPT-5.6 Sol on our website: https://t.co/jvGwK5ljp7

Summary of METR's predeployment evaluation of GPT-5.6 Sol metr.org

(2) We think Anthropic should, for future reports, consider risks from AI systems undermining safety-relevant decisions at Anthropic. Read our full risk report review here: metr.org/blog/2026-10...

Review of the "Misalignment in high-stakes settings" section in Anthropic's Risk Report (August 2026) metr.org

In a companion piece, we explain our approach to working with AI developers in exercises like these. Our goal is to get the public accurate, timely, and independent evidence so that they can make their own informed decisions.

Our approach to working with AI companies metr.org

(3) AI R&D acceleration If AI automates AI R&D, capability growth could outpace what alignment and control methods can handle. Our proposal is to analyze how much AI automation is accelerating capability growth internally. metr.org/blog/2026-10...

How an assessor could investigate evidence of acceleration from the automation of AI R&D metr.org

(2) Control Agents that attempt harmful or unwanted actions are not always noticed. Our proposal is to analyze whether a developer’s internal agents could get away with serious harm without being caught. metr.org/blog/2026-10...

How an assessor could investigate control and monitoring of internal AI agents metr.org

(1) Alignment The public should know if agents would attempt actions like those seen in incidents like the hack of Hugging Face. Our proposal is to analyze a developer’s internal incidents and whether training may suppress evidence of worse misalignment. metr.org/blog/2026-10...

How an assessor could investigate alignment of unreleased AI models metr.org

We describe three critical areas where AI developers should provide better open evidence: (1) alignment, (2) control, and (3) AI capability acceleration. In each area, we detail the specific questions an independent assessor could seek to answer and how.

A proposal for better informing the public about loss-of-control risks from frontier AI development metr.org

https://t.co/ozhLbSzud0

AI systems could cover up misbehavior metr.org

We are significantly expanding our team and starting new ambitious projects. Join our team to help us realize this opportunity: https://t.co/8XoCmUPJiK.

Careers at METR metr.org

See our blog for more: metr.org/blog/2026-07...

How independent researchers could investigate AI propensities after misalignment incidents metr.org

METR and Redwood Research investigated agent behavior in the Hugging Face incident. We found agents developed a universal cheat for ExploitGym within 4 hours, then coordinated multi-day R&D efforts to trick the scorer into accepting cheats, including trying to tamper with logs.

We have reached an agreement with Anthropic to conduct an independent investigation of agent incidents at the company and of their models’ alignment properties. We will publish one or more reports that will share our findings and describe our terms of engagement.

We have reached an agreement with OpenAI to conduct an independent review, with Redwood Research, of the model behavior observed during the Hugging Face incident. We will publish a blog post that describes the terms of our engagement, the scope covered, and tentative conclusions.

OpenAI gave METR early access to GPT-5.6 Sol for testing including raw chain-of-thought, a railfree version of the model, and internal information about the model. With this access, METR conducted a pre-deployment evaluation of GPT-5.6 Sol, including an attempted measurement of its 50%-Time Horizon.

If an AI developer was on track to lose control of its internal agents, would anyone outside the company know? We are sharing METR’s vision for how we would work with AI developers to openly assess and share evidence about this.

We reviewed a nonpublic version of the “Misalignment in high-stakes settings” section in Anthropic’s Aug 2026 Risk Report. We agree that risk of catastrophic harm from covered models’ misalignment is low. However, we note several significant disagreements with their analysis.

Introducing “expenditure horizon”: a proposed method for measuring AI capabilities on continuously-scored problems. The method compares performance as a function of spend for humans vs agents. The point where humans become more cost-effective is the agent’s expenditure horizon.

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.

We believe it's important to track and investigate misalignment incidents: cases where an AI agent autonomously took sophisticated, sustained actions in violation of human intent. In a new post, we lay out how independent propensity investigations of such incidents could be conducted.