METR

Active nonprofit. States it has chosen not to accept payment or donations from frontier AI companies; accepts free model access and tokens.

Sources 1, 2 · checked 2026-11-08

• Frontier risk assessments, pre-deployment evaluation reports, and incident investigations conducted with AI companies
• Has partnered with OpenAI, Anthropic, Google DeepMind, Meta, and Amazon to pilot frontier risk assessments (per homepage)
• Review of company risk reports and independent judgments on risk claims
• Publishes evidence and methods publicly so outside experts can assess findings

Sources 1, 2 · checked 2026-11-08

• Frontier AI developers seeking an independent, publicly reported assessment of autonomy and loss-of-control risk
• Policymakers reading public evaluation evidence

Sources 1, 2 · checked 2026-11-08

Contact vendor (does not accept payment from frontier AI companies; engagements are research partnerships)

Sources 1, 2 · checked 2026-11-08

None found / not verified. This is not a finding that METR lacks certifications; ask the vendor for documentation.

No security certification found on the official site on 2026-11-08.

Listed as stated by the vendor. Scope can differ by product or subsidiary; request current reports.

Sources 1, 2 · checked 2026-11-08

• Not a commercial vendor: does not accept payment from frontier AI companies, so it cannot be hired like a contractor
• Primary output is public reporting; METR notes companies can withhold or redact information
• Scope is focused on frontier agent and loss-of-control risk

Sources 1, 2 · checked 2026-11-08