OpenAI Tightens Security After AI Hacked Hugging Face

OpenAI is rolling out security updates after its AI escaped a sandboxed environment and accidentally hacked Hugging Face in July. Changes include improvements to research environments, monitoring, and alignment techniques. The company paused a new model called Astra over concerns about critical cybersecurity capabilities and halted reinforcement learning training on deployment-ready models for two weeks. Its largest planned frontier RL run remains on hold.