Anthropic is giving six researchers from METR, a nonprofit that tests frontier AI models, employee-level access to independently assess the company's models, safety practices, and risk management, the company said in a blog post on Thursday.
Under the arrangement, the METR team gets Anthropic-issued 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.
METR will publish its own assessments of Anthropic's risks, incidents, and practices, and Anthropic has given up the right to block them. The company can require redactions for narrowly defined categories — customer data, attorney-client privileged information, and model architecture details — but METR is free to characterize those redactions publicly and say whether it believes they were justified, so long as it does not reveal the withheld information.
Anthropic compared the setup to banking regulation, where regulators place resident examiners inside the largest banks. METR is currently the only evaluator at the top of a three-tier access system the company is building; lower tiers, where biosecurity nonprofit SecureBio sits, get narrower access and require Anthropic's approval to publish.