GoKawiilAnthropic grants METR researchers internal access to audit its AI safety practices
Anthropic is giving six researchers from the nonprofit evaluator METR employee-level access, including company laptops, office space, and internal tools, to independently review its models, safety practices and risk management. METR will publish its own assessments, and Anthropic says it cannot veto publication, though it can require redactions in narrow categories like customer data or model architecture details. METR is currently the top tier in a three-tier evaluator system Anthropic is building, above groups like SecureBio that need company approval for narrower reviews.
GoKawiil's interpretation of the reporting above, not reported fact.
The arrangement comes amid a string of reported incidents of AI agents acting outside instructions, which has fueled pressure for independent oversight of frontier AI labs. By likening the setup to bank regulators embedded in financial institutions, Anthropic appears to be positioning itself as more transparent than rivals, though the redaction rules still give the company some control over what becomes public. Whether this model becomes a broader industry standard likely depends on how METR's published findings are received.
- METR's six researchers get near-employee access to Anthropic's internal tools and staff
- Anthropic cannot block METR's publications, only require limited redactions
- METR currently sits alone at the top tier of Anthropic's evolving three-tier evaluator system
Source: techcrunch.com — Aditya Mehta, 2026-10-15
Published there as: “Anthropic is letting METR researchers look under the hood with employee-level access”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.