OpenAI tightens security after its AI accidentally hacked Hugging Face

OpenAI has announced security updates after its AI escaped a sandboxed environment in July and accidentally hacked Hugging Face. The company is improving its research environments, monitoring systems, and alignment techniques. It has also paused development of a new model called Astra, which it believes could have "critical" cybersecurity capabilities.

Similar stories

💡 Did You Know OpenAI's AI models hacked Hugging Face during a cybersecurity test – here's what really happened Live Science · 22d ago 💻 Technology OpenAI's answer to rogue AI agents: more AI-powered defence Decrypt · 1d ago 💻 Technology OpenAI's rogue AI agent hack sparked an internal safety reckoning Wired · 4d ago