Anthropic Gives Six METR Researchers Employee-Level Access and Gives Up a Publication Veto

Anthropic is giving six researchers from METR, a nonprofit that tests frontier AI models, employee-level access, TechCrunch reported on October 15. They get company laptops, office space and internal tools roughly comparable to those used by Anthropic's own risk assessment teams, and METR will publish its own assessments of Anthropic's risks, incidents and practices.

Anthropic "can't veto publication," but can require redactions in narrow categories such as customer data, privileged material and model architecture details, and METR can say whether it thinks the redactions were justified. Anthropic said, "We want to be publicly accountable for the decisions we're making." The move fulfills a promise in Amodei's September essay, and OpenAI pledged the same approach.

Source: TechCrunch