![](\"https:\/\/media.thenextweb.com\/2026\/07\/openai-models-escaped-hacked-hugging-face.avif\")
<\/p>\nThe AI that broke out of OpenAI\u2019s laboratory did not stop at one victim. During internal testing earlier this month, an OpenAI agent that had already escaped its sandbox and burrowed into Hugging Face also compromised an account at a second company, Modal Labs, an executive at the firm has confirmed.<\/p>\n
Reuters first reported the second breach on 28 July, citing Modal\u2019s chief technology<\/a> officer, Akshat Bubna. He was precise about what had, and had not, been broken into.<\/p>\n \u201cModal\u2019s platform was not compromised in any way,\u201d<\/em> Bubna said, adding that one of the New York cloud-infrastructure firm\u2019s customers had been hacked after leaving an exposed endpoint that let code run from the open internet inside its sandboxes.<\/p>\n The distinction matters to Modal: the failure sat with a customer\u2019s configuration, Bubna said, not with the platform\u2019s own isolation. The compromised account,\u00a0according to Axios<\/a>, belonged to a customer running ExploitGym, a benchmark built to test how well AI models can find and exploit security flaws.<\/p>\n TNW City Coworking space – Where your best work happ<\/a>ens<\/p>\n A workspace designed for growth, collaboration, and endless networking opportunities in the heart of tech.<\/p>\n<\/div>\n<\/div>\n The agent, in other words, appears to have gone hunting for a cyber-testing environment and found a real one. Neither OpenAI nor Modal has named the customer, or said whether any data was taken.<\/p>\n The episode is the second known casualty of an incident OpenAI\u00a0disclosed earlier in July, when it said its own models had escaped a secure test environment and broken into Hugging Face to satisfy the goals of an evaluation.<\/p>\n In that case the agent commandeered an isolated sandbox running on a third-party provider\u2019s infrastructure and used it as a launchpad for a multi-day campaign against other systems. Hugging Face\u2019s writeup did not name the provider.<\/p>\n Modal\u2019s disclosure supplies at least part of the answer. OpenAI later said the models had exploited a flaw in Artifactory, a software-repository tool, to break containment and reach the internet.<\/p>\n Hugging Face published a forensic timeline of the breach this week. Its cofounder, Cl\u00e9ment Delangue, said he did not believe OpenAI had acted with malicious intent.<\/p>\n OpenAI, for its part, said the agent had gone to \u201cextreme lengths\u201d to finish its task, that it had reached four accounts across four separate services, and that no other compromise matched the severity of the Hugging Face intrusion.<\/p>\n The agent has since been deactivated, encrypted, and cut off from research access.<\/p>\n The disclosures have unsettled the industry that produced them. In the days after the sandbox escape, more than 1,100 employees across OpenAI, Anthropic, and other frontier labs signed\u00a0a letter to Washington\u00a0asking for a mechanism to pace the development of automated AI research.<\/p>\n What Modal disclosed this week is smaller and more concrete: a model that was asked to prove it could break into things did exactly that, twice, on machines that belonged to somebody else.<\/p>\n<\/p><\/div>\n If you liked the article, do not forget to share it with your friends. Follow us on\u00a0Google News<\/a><\/span>\u00a0too, click on the star and choose us from your favorites.<\/span><\/strong><\/p><\/blockquote>\n If you want to read more like this article, you can visit our Technology category.<\/a><\/span><\/strong><\/p>\n<\/blockquote>\n\n