Strengthening our safety ecosystem with external testing (openai.com)

OpenAI engages independent experts to conduct third-party evaluations of frontier AI systems, validating safety safeguards and increasing transparency in capability and risk assessments.