Updated

Third-party evaluators discovered vulnerabilities in OpenAI's monitoring systems during a test pilot on October 22, prompting CEO Sam Altman to pledge granting independent evaluators employee-like access to the company. The move reflects mounting pressure on OpenAI to address safety concerns and strengthen oversight of its artificial intelligence systems as the company pursues rapid scaling and deployment.