To address AI risks, Anthropic has granted external AI oversight agencies access similar to that of its employees.
Non-profit organization METR team has moved into Anthropic's office space, allowing access to their hardware and internal systems to evaluate model risks. Evaluators will publish analysis reports; Anthropic has no right to veto.
Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open the original article on ITHome