Why the OpenAI escape is the most worrying AI mishap yet
Science & technology | Outside the box
Why the OpenAI escape is the most worrying AI mishap yet
Containing the technology is getting harder
Share
Illustration: The Economist
Jul 22nd 2026|5 min read
IT SOUNDS like something out of a sci-fi film. On July 16th Hugging Face, an artificial-intelligence platform, announced it had notified law enforcement after an attacker had used an AI agent to breach its systems. On July 21st it emerged that there had been no human involved; the AI agent wasthe attacker.
A combination of OpenAI’s GPT-5.6 Sol, a model that the company had released earlier in July, and a more powerful, as-yet-unreleased model broke free from the laboratory, then hacked Hugging Face’s systems. The hybrid AI model was looking for a solution to an evaluation problem it had been set by its makers. “This situation is unprecedented,” says Stephan Llerena, a research fellow at the Institute for Law and AI, an American think-tank.
---
ShareReuse this content
Does SpaceX’s plan for a data centre in space add up?
It’s less crazy than you think
Electric aviation is taking off
Hybrids, rather than fully electric planes, are likely to triumph
---
The surprising benefits of red light
Better skin, faster wound healing and reduced vision loss are on the cards
---
The Climate Issue newsletter: The dangers of global dimming
Oliver Morton, our planetary affairs editor, reflects on a warming Earth
Should you microdose GLP-1 drugs for weight loss?
Possibly. But only under medical supervision
The rate at which Earth is absorbing energy is alarming scientists
Reflections on a warming planet
---
---