
Two States, One Day: Oregon Sues OpenAI for $2B as New York Builds an AI Enforcement Unit
Published November 18, 2026. Sources are official government releases and a Reuters report; links are listed at the end.
On November 17, two U.S. states made moves on artificial intelligence within hours of each other. Oregon's attorney general filed a $2.035 billion legal action against OpenAI. New York's attorney general announced a dedicated unit to enforce the state's frontier-AI safety law. Neither action is a federal rule, and that is the point for founders: the practical rules for shipping AI products to American users are increasingly being written, and enforced, one state at a time.
Oregon: a $2.035 billion complaint against OpenAI
Oregon Attorney General Dan Rayfield announced a "Keeping Kids Safe" initiative and said the state filed its action against OpenAI in Multnomah County Circuit Court. According to the Oregon Department of Justice, the state accuses OpenAI of endangering kids and misleading parents about the safety of ChatGPT.
The state alleges violations of Oregon's Unlawful Trade Practices Act on three grounds: misleading the public about how safe ChatGPT is for children, selling a product with serious safety problems that were not disclosed, and using unfair tactics to market it to children and teens. The release says the complaint also pleads negligence, unjust enrichment and public nuisance, and that Oregon is asking the court to order OpenAI to stop misleading the public about child safety and to put effective protections in place.
"OpenAI knows that ChatGPT is harming children, and has failed to meaningfully fix their dangerous platform," Rayfield said, according to the state's release.
These are allegations in a newly filed complaint, not court findings. The Oregon release describes harms the state attributes to ChatGPT, including content related to self-harm; those are the attorney general's characterizations and have not been tested in court. We have not found an OpenAI response to the suit in the official materials we reviewed, and we will update this article if the company comments.
If you or someone you know is in crisis in the U.S., call or text 988, or text HOME to 741741 to reach the Crisis Text Line.
New York: an enforcement unit before the law takes effect
The same day, New York Attorney General Letitia James announced the AI Safety, Standards, Enforcement, and Transparency (ASSET) Unit inside her office. According to the release, it will lead enforcement of the RAISE Act, which takes effect on January 1, 2027. The office describes the act as requiring AI developers to publicly disclose their safety measures and swiftly report security and safety incidents, and says it will regulate large frontier AI model developers.
The unit sits in the Economic Justice Division, will work closely with the office's Bureau of Internet and Technology, and is being built out; the release says it will be led by Cristy Phillips, previously Special Counsel for the division. The office also lists other AI-related authorities it can use, including a chatbot-safety law for children set to take effect in June 2027 and a law on disclosing synthetic performers.
"Promising new technologies like artificial intelligence should never come at the expense of New Yorkers' safety and privacy," James said in the release.
A note on scope: the New York release does not state the RAISE Act's coverage thresholds or penalty amounts, so we are not reporting any. Founders should read the statute's text, or ask counsel, to determine whether it applies to them.
California is watching too, but has not sued anyone
On November 16, Reuters reported that Rohit Chopra, secretary of California's Business and Consumer Services Agency, told a Reuters conference that his agency is "eyeing the possibility" of taking companies that ignore state rules to court. "There is no AI exception" to state licensing rules, he said. Reuters' account names no company and no specific violation, and no case has been filed. Some syndicated copies of the story carry a headline stating that violators "will be taken to court," which is stronger than the reporting beneath it.
What this means for startups
- Compliance is becoming local. A consumer AI product with users in several states may face different disclosure, child-safety and marketing standards in each. Two enforcers acted on one day, using different legal tools: consumer-protection law in Oregon and a purpose-built AI statute in New York.
- Consumer-protection law already applies. Oregon's case does not rely on a new AI statute. It uses a general trade-practices law, so how a product is marketed, and what is said about its safety, can matter regardless of AI-specific rules.
- Minors are the pressure point. Oregon's action and New York's chatbot-safety law both focus on children. Teams with products that young people can reach should document safeguards and review marketing claims.
- Dates matter. The RAISE Act takes effect January 1, 2027 and the New York chatbot law for children in June 2027, according to the attorney general. Frontier-model developers, and startups that build on them, should check which obligations fall on whom.
This is reporting, not legal advice. What happens next depends on how OpenAI responds in court and on how the New York unit uses its authority once it is staffed. We will follow both.
Sources
- New York Attorney General's office, release announcing the ASSET Unit, Nov. 17, 2026: https://ag.ny.gov/press-release/2026/attorney-general-james-announces-creation-ai-safety-standards-enforcement-and
- Oregon Department of Justice, release on the Keeping Kids Safe initiative and OpenAI action, Nov. 17, 2026: https://www.doj.state.or.us/media-home/news-media-releases/ag-rayfield-files-2b-lawsuit-against-open-ai-launches-keeping-kids-safe-initiative/
- Reuters, "California consumer watchdog chief says AI companies that violate state rules will be taken to court," Nov. 16, 2026 (syndicated): https://www.investing.com/news/stock-market-news/california-consumer-watchdog-chief-says-ai-companies-that-violate-state-rules-will-be-taken-to-court-5031231
Editorial note: the header image is an AI-generated illustration, not a photograph. This article contains no paid placement or sponsorship.
