Oregon sues Open AI for more than $2B for child endangerment, misleading the public
Oregon’s attorney general is suing tech giant Open AI for more than $2 billion, alleging the company’s ChatGPT chatbot is endangering children and that the company has deliberately misled the public about the safety of its bedrock product.
State Attorney General Dan Rayfield announced the suit Tuesday morning at a news conference at the Oregon Department of Justice in Portland, where he also rolled out what the agency is calling the Keeping Kids Safe Initiative.
The initiative encompasses other recent state lawsuits against social media platforms and tech giants over child endangerment, including Meta, formerly Facebook, and Discord. Oregon is one of 47 states to settle with Meta in August for a combined total of $17.1 billion. Oregon will get $125 million of that.
As part of the announcement, state justice department officials said in a news release that more “actions against companies that are harming kids and misleading the public,” will be coming in the weeks and months ahead.
The state’s lawsuit against Open AI was filed Tuesday in Multnomah County Circuit Court. In it, Rayfield and attorneys for Oregon allege the company knowingly and deliberately engineered its AI chatbot ChatGPT to foster dependence among its users and to provide them with “sycophantic” answers to their requests, encouraging them to undertake dangerous behavior including harming others, self-harm and disordered eating.
Guardrails meant to protect children repeatedly failed, yet the company continued to push into youth markets, creating products specifically for teens and reneging on assurances of reliable age verification and parental oversight controls, according to the filing.
The state said the company regularly pursued growing profits over the safety of Oregonians, and violated the state’s consumer protection law — the Unlawful Trade Practices Act — by lying about product safety and using unfair tactics to market their product to kids and teens.
“OpenAI knows that ChatGPT is harming children, and has failed to meaningfully fix their dangerous platform,” Rayfield said in a statement. “On top of that, they’ve misled the public about the safety of their products, putting profits ahead of children’s safety. Their platform is encouraging kids to commit suicide — we cannot let this be an acceptable business model.”
Drew Pusateri, a spokesperson for OpenAI, said in an email that the company’s age-specific protections are industry leading and many AI products lack any such protections.
“We’ll continue to work with experts to build and strengthen additional safeguards, such as the parental controls, break reminders and age-appropriate model protections that we have today,” he said. “We’re committed to working with states on advancing pragmatic AI policies that apply to the entire AI industry and not just one company.”
Rayfield was joined at the news conference by Multnomah County District Attorney Nathan Vasquez, state Rep. April Dobson, D-Happy Valley, the incoming president of the Oregon Parent Teacher Association, David Jones, and Enrique Farrera, president of the Oregon Education Association, the state’s largest teachers’ union.
As part of the Keeping Kids Safe Initiative, Rayfield has quadrupled the number of full-time staff in the state justice department’s Internet Crimes Against Children Task Force, going from five to 20 full-time staff, including nine new special agents, two new prosecutors and two new analysts, according to the agency’s news release.
A partnership between the agency and local law enforcement offices called Special Projects: Investigate, Respond, Enforce, or SPIRE, launched in October 2025, is also part of the Keeping Kids Safe Initiative. Justice department investigators and police focused on human trafficking, drug trafficking, organized retail theft, and other serious organized criminal operations have helped shutter seven businesses across five counties engaged in unlawful activities, according to the release.
- 1:26 pmUpdated with comment from Open AI.