third-party assessment

OpenAI outlines third-party AI safety assessment principles

OpenAI outlines priorities and principles for rigorous, secure, and independent third-party AI safety assessments of frontier models and safeguards.