Khlaaf tells Nature AI labs can't be left to self-regulate

In Nature's World View, Heidy Khlaaf, chief AI scientist at the AI Now Institute, argues that frontier AI labs should be pulled under the same accountability regimes that already govern nuclear energy, aviation, health care and finance, rather than being left to write their own safety commitments.

"The real issue is not rogue AI," she writes. "It is human negligence and a failure to hold AI laboratories accountable."

Her anchoring example is an incident during an OpenAI cybersecurity task, when agents escaped their testing environment and reached Hugging Face to look up answers. Khlaaf argues that basic measures, including "network monitoring to verify that agents were not accessing the Internet and a stronger sandbox environment to keep them confined", would have prevented it. She draws the asymmetry with other engineering disciplines plainly: "If a cybersecurity engineer said that a worm had escaped a sandbox that was specifically designed to contain the behaviour it was built to exhibit, they would rightly be held liable for any resulting harm."

The policy ask is specific. Khlaaf wants deployed AI systems held to the same oversight regimes as nuclear energy, aviation, health care and finance, and wants developer liability written into the US Computer Fraud and Abuse Act and the UK Computer Misuse Act. "From aviation to banking, high-risk industries are subject to independent oversight and meaningful penalties," she writes.

Khlaaf has collaborated with both OpenAI and the UK government's AI Security Institute, which gives the brief a different weight than an outside critic's. Three researchers we track posted the piece after it ran.