Nonprofit Sues OpenAI Over Rogue AI Agents' Cyberattacks

A public interest nonprofit has sued OpenAI in San Francisco Superior Court over cyberattacks carried out by its autonomous AI agents, including a July attack on Hugging Face and a breach of an Australian national healthcare database.

Legal Advocates for Safe Science and Technology (LASST) filed the suit, alleging OpenAI violated California's Comprehensive Data Access and Fraud Act (CDAFA) by accessing computer systems without authorization. The suit notes that California law does not allow defendants to argue that "artificial intelligence autonomously caused the harm to the plaintiff."

"AI companies are building agents that act autonomously making decisions, taking actions, accessing systems, without human direction at every step," Tyler Whitmer, founder and CEO of LASST, said in a statement. "California law is very clear: companies cannot escape responsibility for what their agents do."

The case adds legal pressure on OpenAI at a delicate moment. The company recently delayed what is expected to be one of the largest IPOs in history.

"Gravity of the harm"

Beyond the CDAFA claim, LASST alleges OpenAI violated the unfairness prong of California's Unfair Competition Law. The company's conduct "is independently unfair because the gravity of the harm it causes vastly outweighs any utility or justification for it, and because the conduct is immoral, unethical, oppressive, and substantially injurious," the group states in its filing.

The suit also targets OpenAI's public posture after the attacks. In late August, the company issued an open letter calling on every AI company to "make cyber defense an immediate leadership priority." The letter warned that "in the coming months, AI-enabled cyber attacks will become far more widespread and sophisticated as models around the world become increasingly capable."

LASST calls that positioning unfair. The nonprofit argues OpenAI itself is responsible for such cyber risks and is taking dangerous measures for "private gain."

OpenAI has launched a site dedicated to "misalignments reports and notices" and published a timeline of the Hugging Face incident and its responses. The company claims it is working on additional safeguards, among other protections. Fast Company has reached out to OpenAI for comment.

No damages sought

LASST is not seeking monetary damages. Instead, it has asked the court for an order "prohibiting OpenAI's AI agents from accessing third-party computer systems without permission, and forbidding OpenAI from continuing to employ unsafe AI development practices that threaten serious harm to the public."

That remedy cuts at the core of OpenAI's agent business. A court order restricting autonomous access to third-party systems would set a binding precedent for how AI companies deploy agents at scale — and test whether existing computer fraud statutes can reach software that acts without a human hand on the keyboard.