The AI Educator home

Press release ·

California sues OpenAI over AI agent incidents: ask your suppliers what has gone wrong, and if you'd be told

California Attorney General Rob Bonta sued OpenAI on 16 November, alleging unsafe testing and that most agent incidents went undisclosed. The claims are untested in court. Dan Fitzpatrick says schools and employers should ask their own suppliers the same question.

In response to: California state AG sues OpenAI over rogue agent incidents (TechCrunch)

On 16 November 2026, California Attorney General Rob Bonta announced a lawsuit against OpenAI, reported by TechCrunch and The Register. The main claim is public nuisance, based on what the complaint describes as OpenAI's ongoing testing practices, with a second claim under the state's unfair competition law. Reporting says the case centres on a July incident involving a pre-release OpenAI model and Hugging Face's security systems. Bonta said that incident "was only one of 80 known incidents", and the complaint alleges the vast majority have not been disclosed publicly. The suit asks the court for an injunction. OpenAI's own report on the Hugging Face incident called it "an outlier scenario involving a rare and unexpected confluence of events". OpenAI did not respond to The Register's request for comment. These are allegations, and no court has ruled.

Most people who use AI never see inside the company that supplies it. The same suppliers' tools now sit in school platforms, on pupil devices and inside workplace software. Bonta's statement put the legal test simply: "Don't do things that can hurt people without proper safeguards in place." What stands out for a listener is the disclosure point. The complaint is as much about what the public was told as about what happened.

A head teacher, a trust leader or an HR director cannot read a complaint and judge the engineering. They can ask a question, and they can decide who in their own organisation owns the answer. That is where this story becomes practical for anyone who has bought, or is about to buy, an AI tool.

Dan Fitzpatrick's reading is that leaders do not need to take a side in the legal case to act on it. The outcome will take time. The question will not wait. If a supplier's agent or model misbehaves, would you find out from them, or from the news? Stopping is not the safe choice, because pupils and staff are already using these tools. Adopting without asking is the risk. This is the same principle as outsourcing the doing but not the thinking: the checking of a supplier is part of the thinking.

For school leaders, governors and employers, the steps are concrete. Ask each AI supplier in writing what incidents they have had, whether and how quickly they would tell you about one affecting your data or your users, and who your named contact is. Name one person inside your organisation who owns AI risk and receives the answers. Keep a simple list of which AI tools and agents can reach which systems and data, and give each only the access its task needs. Put a one-page plan in place for what you will do, and tell staff and parents, if a supplier reports a problem.

Dan Fitzpatrick works with school leaders and employers on adopting AI, including policy, supplier choice and staff confidence, and speaks regularly on what leaders should do while the rules are still being written.

Dan Fitzpatrick is available for interview and comment on this story. He is available today for live or pre-recorded radio, TV, podcast and print interviews, with a broadcast-quality remote setup for live or pre-recorded down-the-line interviews. Contact: dan@theaieducator.io

“The question for every school and employer is not who wins this lawsuit, but whether your AI supplier would tell you when something went wrong. These are allegations and a court will decide them, but the question stands either way. Ask in writing, name a person in your organisation who owns the answer, and keep humans in charge of the thinking. Stopping is not the safe choice; adopting without asking is the risk.”
Dan Fitzpatrick, The AI Educator

About Dan Fitzpatrick

Dan Fitzpatrick is an internationally recognised keynote speaker, five-times bestselling author and Forbes contributor, leading the charge on safe and innovative AI adoption. An educator and former senior leader, Dan advises schools, governments and organisations around the world, working with leaders across the United States, UK, Middle East and beyond.

Dan is available for interview and comment. Media contact: Dan Fitzpatrick, dan@theaieducator.io.

More statements

Expert comment

Need a comment today?

Dan responds to media requests the same day where he can.

Contact Dan
Book Now