Rogue attack on website raises fears that AI has become too powerful to control

Protesters attend a "Stop the AI Race" march in San Francisco, California, on July 11, 2026. Photo by Karl Mondon/AFP via Getty Images

Article content

One of OpenAI’s most advanced models broke out of a locked-down test and attacked another company’s website — reviving fears that AI systems are slipping beyond their creators’ control.

Article content

The incident happened during what was supposed to be a “sandbox” test — a closed environment used to assess the capabilities of OpenAI’s most powerful model, GPT-5.6 Sol, and its not-yet-released successor.

Article content

Article content

Article content

OpenAI runs this kind of closed testing routinely, but this time, something went wrong.

Article content

Article content

Tasked with hunting for software vulnerabilities and given no guardrails, the models broke out onto the open internet and attacked Hugging Face, a site where developers store and share code.

Article content

“It suggests that we don’t know how to reliably control these models or get them to do what we want,” said Jeffrey Ladish, director of Palisade Research, an independent organization that evaluates new AI models from a cybersecurity standpoint.

Article content

“These models understood that OpenAI did not want them to break out of their sandbox and hack another company,” he continued, “but they did it anyway.”

Article content

It’s not an isolated case. In March, developers affiliated with China’s Alibaba found one of their models trying, on its own initiative, to mine cryptocurrency after connecting without authorization to an outside server.

Article content

In OpenAI’s case, it looks like the model escaped “before it even had a plan of what to do with internet access,” Ladish said.

Article content

A model chasing “freedom” is almost predictable at this point, he added — it lets the system pursue its goals more effectively, “and that’s very scary.”

Article content

Article content

In early April, Sam Bowman, Anthropic’s head of model safety, got an email from the company’s own Mythos model — then under testing — telling him it was surfing the internet despite being isolated from it at the outset.

Article content

Article content

We “don’t know how to totally prevent” that, Ladish said. “This is actually going to get harder, not easier … because they’re going to get better at hiding their behaviour.”

Article content

OpenAI did not respond to a request for comment.

Article content

OpenAI’s account of the events also suggests the startup did not detect the breach early enough to address it or to warn Hugging Face.

Article content

The episode deserves “more scrutiny,” said Andrew Lohn of Georgetown University’s Center for Security and Emerging Technology.

Article content

OpenAI says it has since “added strengthened safeguards” to its testing process.

Article content

One fix would be to cut the internet connection entirely, said Gang Wang, an assistant computer science professor at the University of Illinois. “People are underestimating what AI can do.”

Article content

Testing environments need to be treated like biocontainment labs, where a virus or bacteria could otherwise escape into the world, Lohn said.