Anthropic is giving employee-level access to six researchers from METR, a nonprofit that tests frontier AI models, so they can independently assess the company's models, safety practices and risk management. The company announced the arrangement in a blog post on Thursday.
The timing matters. Incidents of AI agents acting outside their instructions have been disclosed every week or two since late August, and those cases have sharpened calls for independent checks on AI companies. Anthropic framed the new setup as a direct response.
"We want to be publicly accountable for the decisions we're making," the company 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 receives Anthropic-issued company laptops, office space and access to internal tools that are roughly on par with what Anthropic's own risk assessment teams use. Anthropic said its employees are encouraged to share information with them "just as they would with a fellow employee working on alignment risk."
The company drew a comparison to banking. Regulators place resident examiners inside the largest banks, and Anthropic said the METR arrangement works on similar principles.
METR will publish its own assessments of Anthropic's risks, incidents and practices. Anthropic has given up the right to block those reports. It does retain a limited power to redact. "Anthropic can't veto publication, but we can require them to redact information if it falls into narrow, pre-defined categories," the company said. Those categories include customer data, anything attorney-client privileged and information about its model architectures.
Anthropic also set terms for how METR can respond to redactions. "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," the company said.
METR is currently the only evaluator at the top of a three-tier system Anthropic is building. Evaluators in the bottom tier, such as the biosecurity nonprofit SecureBio, run narrow reviews and need Anthropic's approval before publishing. The middle tier receives more information and can publish without approval. The top tier adds METR's employee-level access.
The announcement carries out a promise Amodei made in a September essay. In that essay, he proposed embedding outside evaluators inside AI labs as a first step toward slowing the technology's development. OpenAI quickly pledged to do the same. Anthropic, OpenAI, Google, Meta, Nvidia and xAI have since signed a White House accord committing them to independent review.
Dieser Artikel wurde mit Unterstützung von KI verfasst.
News Factory APP - agentische News für besseres SEO & AEO.