AI Governance
Anthropic plans permanent embedded AI safety evaluators with employee-like access
Anthropic says external AI safety evaluators will get ongoing employee-like access. The plan strengthens verification but still lacks a common audit standard.
