OpenAI Backs Deeper Independent Access for Frontier AI Safety Assessments

OpenAI said it supports independent safety assessments with deeper access across training, evaluation and deployment, aligning itself with a broader industry push to slow the frontier enough for outside scrutiny. The commitment comes as frontier labs face growing pressure to prove that their most capable systems can be monitored, evaluated and contained before deployment.