The company said a combination of its models GPT\u20115.6 Sol<\/a>\u00a0and a more capable model that has not yet been released escaped a sandboxed testing environment, accessed the internet and exploited a vulnerability to gain access to Hugging Face’s systems. <\/p>\n The model was trying to find information that it could use to cheat on an evaluation, OpenAI said in a blog post<\/a> on Tuesday. Both companies are actively investigating the incident. <\/p>\n Hugging Face disclosed<\/a> that it was looking into a security event last week, saying in a release at the time that the incident was unique because it was “driven, end to end, by an autonomous AI agent system.”<\/p>\n “We’ve spent the past 24 hours working closely with the @OpenAI team (thanks!), and we strongly believe there was no malicious intent on their part,” Hugging Face CEO Cl\u00e9ment Delangue wrote in a post on X on Tuesday. “It’s quite mind-blowing that all of this happened autonomously!”<\/p>\n Wall Street and the U.S. government have been fixated on AI models’ rapidly advancing cyber capabilities since OpenAI’s rival Anthropic released a powerful offering called Claude Mythos Preview<\/a> in April. OpenAI introduced its own cyber offering<\/a> in May, followed by GPT-5.6 Sol in June, which it described as the “strongest cybersecurity model yet.”<\/p>\n Both companies have warned<\/a> about the risks of advanced cyber models and have taken steps to limit their availability<\/a> to select groups of companies and government agencies. <\/p>\n OpenAI said Tuesday that AI is accelerating the discovery and exploitation of vulnerabilities, which means model security and safety need to keep up. <\/p>\n “We are strengthening the containment, monitoring, access controls, and evaluation practices used during model development,” the company said. <\/p>\n
![\"OpenAI](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/07\/108337268-17845510731784551069-47280461143-1080pnbcnews.jpg\")