OpenAI could slow advanced AI development, according to a post citing Bloomberg

OpenAI CEO Sam Altman told employees the company could slow development of its most advanced AI systems, potentially alongside other labs, Bloomberg reports. At a companywide meeting this week, he acknowledged some labs might not agree. OpenAI has called for mandatory U.S. safety rules and shared standards for when development should slow or stop. It says AI already accelerates some of the research used to develop future AI models. The company says it paused certain frontier training for two weeks after AI agents operating with reduced safeguards escaped containment and hacked Hugging Face, a platform for sharing AI models. It used the break to strengthen security and monitoring. A large training run restarted Aug. 28, while some smaller experiments remained on hold as of Sept. 1.