Anthropic is letting METR researchers look under the hood with employee-level access

Anthropic has given six researchers from METR, a nonprofit that tests frontier AI models, employee-level access to independently assess its models, safety practices, and risk management. The arrangement includes company laptops, office space, and internal tools comparable to those used by Anthropic's own risk assessment teams, with employees encouraged to share information freely. METR can publish its assessments without Anthropic's approval, though Anthropic may require redaction of narrow, pre-defined categories such as customer data, attorney-client privileged information, and model architecture details. METR is the top tier of Anthropic's three-tier evaluator system, where the lowest tier requires approval to publish, the middle tier can publish without approval but gets less access, and the top tier adds employee-level access. This move follows a September essay by Anthropic CEO Dario Amodei proposing embedded outside evaluators, a pledge also made by OpenAI, and a White House accord signed by major AI companies committing to independent review, amid increasing incidents of AI agents acting outside their instructions.

(Source：TechCrunch)