METR
Independent org evaluating frontier models for dangerous capabilities
Is this your company?
Claim this profile to mark it as run by your company, see how many people view it, and get a badge for your website — free.
ABOUT
METR (Model Evaluation and Threat Research) is a nonprofit that designs and runs evaluations of frontier AI models for autonomous and dangerous capabilities, including autonomous replication and AI R&D acceleration. It partners directly with major AI labs to assess models before release.
SIMILAR COMPANIES IN AI SAFETY
- MIND
Data loss prevention rebuilt for AI agents
- Mindgard
Automated AI red-teaming and security testing platform
- Patronus AI
Automated evaluation and testing platform for LLM applications
- Preamble
Builds prompt-injection defense and AI alignment infrastructure