Anthropic gives METR employee-level access to scrutinize its models

Anthropic is giving employee-level access to six researchers from METR, a nonprofit that tests frontier AI models, so they can independently assess Anthropic’s models, safety practices, and risk management, the company said in a blog post on Thursday. The arrangement comes as incidents of AI agents acting outside their instructions have been disclosed every week or two since late August, incidents that have sharpened calls for independent checks on AI companies. “We want to be publicly accountable for the decisions we’re making,” Anthropic said in the post. “We aim to close the gap between what evaluators understand about a frontier AI company’s systems and what the company itself understands.”

Under the arrangement, the METR team has Anthropic-issued company laptops, office space, and access to internal tools roughly on par with what Anthropic’s own risk-assessment teams use. “Anthropic employees are encouraged to share information with them just as they would with a fellow employee working on alignment risk,” the company said. Anthropic compared the setup to banking regulation, likening it to how regulators place resident examiners inside the largest banks.

METR will publish its own assessments of Anthropic’s risks, incidents, and practices, and Anthropic has given up the right to block that publication. “Anthropic can’t veto publication, but we can require them to redact information if it falls into narrow, pre-defined categories,” the company said, citing examples such as customer data, attorney-client-privileged material, or information about its model architectures. “METR is free to characterize our redactions publicly, and say whether it believes the redactions were justified, as long as it does so without revealing the redacted information,” Anthropic added.

METR is the only evaluator so far at the top of a three-tier system Anthropic is building. The bottom tier, which includes biosecurity nonprofit SecureBio, runs narrow reviews that require Anthropic’s approval before publication; the middle tier gets more information and can publish without approval; the top tier adds METR’s employee-level access. Thursday’s announcement makes good on a promise Anthropic CEO Dario Amodei made in a September essay proposing that outside evaluators be embedded inside AI labs as a first step toward slowing the technology’s development. OpenAI quickly pledged the same, and Anthropic, OpenAI, Google, Meta, Nvidia, and xAI have since signed a White House accord committing to independent review.

Why it matters

Giving an outside nonprofit standing access to internal tools and the right to publish without a veto is a notably deep concession for a frontier AI lab, and it sets a bar other signatories of the White House accord will now be measured against as they build out their own evaluator tiers.