Anthropic integrates external evaluators to mitigate AI risks

Anthropic has granted employee-like access to external evaluators to analyze the risks of its artificial intelligence models. METR, a non-profit organization...
Read on GoYou.it (goyou.it/en/tecnologia/2026/10/19/anthropic-in…)