GoKawiilAnthropic grants METR auditors 'employee-like' access to internal AI systems
Anthropic said it will give six evaluators from third-party nonprofit METR hardware, office space and internal system access comparable to its own risk-assessment teams, excluding items like customer data or model architecture details. METR will independently publish its safety findings on Anthropic's models, and Anthropic says it cannot withhold any of those conclusions, though it may request redactions of sensitive material.
GoKawiil's interpretation of the reporting above, not reported fact.
The arrangement, first floated by CEO Dario Amodei in a September blog post, suggests Anthropic is trying to demonstrate voluntary safety accountability as AI labs face growing scrutiny over security incidents and as no legal requirement yet mandates third-party testing of frontier models. If other labs follow suit, this could become an informal industry standard for independent oversight before regulation catches up.
- Anthropic is giving six METR evaluators near-employee-level access to internal tools and systems.
- METR will publish its safety findings independently, with Anthropic unable to block the results, only request redactions.
- The move comes amid rising regulatory pressure on AI labs, despite no current legal mandate for third-party audits.
Source: zdnet.com — Radhika Rajkumar, 2026-10-16
Published there as: “Anthropic gives third-party auditors ’employee-like access’ in AI industry first”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.