Anthropic has granted METR researchers employee-level access to its systems, allowing independent auditing of its AI safety practices. METR will publish its own risk and incident assessments without Anthropic's ability to block the findings.
Background
METR (Model Evaluation and Transparency Research) is an AI safety research organization focused on evaluating frontier AI models. Anthropic's move reflects growing industry pressure for independent transparency in AI safety claims.
- Source
- TechCrunch
- Published
- Oct 16, 2026 at 03:14 AM
- Score
- 7.0 / 10