OpenAI Mandates Safety Cases for Frontier AI Training and Executive Halt Authority

This policy represents a substantive step in AI governance to ensure safety and accountability in frontier model development.

Auto-scored: The event describes a mitigation-focused action that reduces risk through enforced governance measures.