Model Evaluation & Threat Research

We inform the public about the risks and capabilities of advanced AI systems.

About METR

Featured work

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

Frontier Risk Report (February to March 2026)

Task-Completion Time Horizons of Frontier AI Models

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

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

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

We are Changing our Developer Productivity Experiment Design

Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity

Measuring AI Ability to Complete Long Software Tasks

Red-Teaming OpenAI's Internal Agent Monitoring Systems

Red-Teaming Anthropic's Internal Agent Monitoring Systems

Early work on monitorability evaluations

Inspect Hawk: our open-source platform for running AI agent evaluations at scale

View website

Implementing and Evaluating a Basic Per-Action Monitor for Safer Evals

Measuring autonomous AI capabilities — resource collection

Frontier AI Safety Policies

Common Elements of Frontier AI Safety Policies

Partnerships

We have previously partnered with OpenAI, Anthropic, Google DeepMind, Meta and Amazon to pilot frontier risk assessments. For our current principles for these engagements, see our approach to working with AI companies. These companies have also provided access and tokens, which we use for evaluations, research, and engineering.

We are also part of the NIST AI Safety Institute Consortium and California Cybersecurity Task Force, are partnering with the AI Security Institute, and provide technical assistance to the European AI Office.

METR in the press

View all press coverage

Recent

Red-Teaming OpenAI's Internal Agent Monitoring Systems

October 22, 2026

A METR employee and a Redwood Research employee spent three weeks stress-testing OpenAI's internal misalignment monitoring system, discovering a number of novel vulnerabilities.

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

October 20, 2026

External review from METR of the "Misalignment in high-stakes settings" section in Anthropic's August 2026 Risk Report

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

October 14, 2026

Our proposal is to analyze how much AI automation is accelerating capability growth internally.

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

How an assessor could investigate alignment of unreleased AI models

October 14, 2026

Our proposal is to analyze a developer’s internal incidents and whether training may suppress evidence of worse misalignment.

How an assessor could investigate alignment of unreleased AI models

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

October 14, 2026

The key questions about internal AI development the public needs more evidence about, and a concrete proposal for how developers could work with METR to produce and share that evidence.

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

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

October 14, 2026

Our proposal is to analyze whether a developer’s internal agents could get away with serious harm without being caught.