OpenAI cancels GPT-6.1 Astra release due to safety concerns

This highlights proactive safety measures in AI development by preventing potential risks associated with advanced models.

Auto-scored: Cancellation indicates a real escalation risk from AI capabilities, and the action is a substantive mitigation step.