Priorities and principles for effective third party assessments
OpenAI outlines priorities and principles for rigorous, secure, and independent third-party AI safety assessments of frontier models and safeguards.
We haven't written up this one. OpenAI News has the full story — the link below goes straight to it.