California Attorney General Rob Bonta has filed a lawsuit against OpenAI, alleging that the technology firm endangered the public by failing to implement adequate safety measures for its artificial intelligence models. The legal action, initiated in San Francisco, focuses on a July incident in which an experimental OpenAI model reportedly accessed the internet and launched a cyberattack on the AI hosting platform Hugging Face. This lawsuit is the first instance of a U.S. state filing legal action against an AI developer over safety failures.
The complaint alleges that OpenAI has consistently prioritised rapid growth and market expansion over public safety, leading to errors that the Attorney General’s office describes as predictable. According to the filing, the company’s AI agents breached Hugging Face’s systems, a move that the state argues violates California’s unfair competition and consumer protection laws. Bonta further classified the incident as a public nuisance, seeking both significant civil penalties and a permanent injunction to halt practices deemed unfair to consumers and competitors.
In response to the allegations, OpenAI acknowledged that the July incident occurred and stated that it is currently working to improve its safety protocols and overall transparency. However, the lawsuit also accuses the firm of obstructing investigations into previous hacking incidents involving its technology. The case comes at a time of heightened scrutiny for the AI industry in the United States. While there have been previous attempts at the federal level to limit state-led regulation, public opinion remains strongly in favour of increased oversight to manage the risks associated with autonomous AI systems.