Third-party AI Testing

OpenAI Sets Principles for Independent Frontier AI Safety Assessments

OpenAI sets priorities for rigorous, secure and independent third-party safety assessments of frontier AI models and safeguards, but details remain limited.