Advertisement

Technology · 22 November 2026

Oregon Sues OpenAI for $2 Billion, Accusing ChatGPT of Endangering Children

Oregon Sues OpenAI for $2 Billion, Accusing ChatGPT of Endangering Children
Oregon's attorney general announcing the lawsuit against OpenAI.Image: The Corvallis Advocate

Oregon has sued OpenAI for at least $2.035 billion, accusing the company of endangering children through ChatGPT and misleading parents about how safe the chatbot really is. The lawsuit, filed on November 17 in Multnomah County Circuit Court by Attorney General Dan Rayfield, is the largest state action yet in a rapidly growing wave of litigation over AI chatbots and young users.

Rayfield announced the case in Portland as he launched "Keeping Kids Safe," a new Oregon Department of Justice initiative to hold companies accountable when they harm children and mislead the public about safety. The claims in the lawsuit are allegations that have not yet been tested in court.

What Oregon is claiming

The state brings the case on behalf of Oregon and its residents against six entities in the OpenAI group, which the filing describes as a single enterprise — meaning each could be liable for the full amount. Oregon is seeking damages of at least $2.035 billion, plus civil penalties, restitution, and a court order requiring changes to the company's practices.

Advertisement

The complaint makes five claims. The first two rest on Oregon's Unlawful Trade Practices Act, the state's main consumer protection law: they accuse OpenAI of knowingly misrepresenting ChatGPT's safety, failing to disclose known defects, and using unfair tactics that exploited children and users with mental health vulnerabilities. The remaining claims are negligence, unjust enrichment and public nuisance. Before filing, the attorney general's office says it notified OpenAI and gave it a chance to commit to changing its conduct, but no satisfactory commitment was offered. Oregon has requested a jury trial.

The design allegations at the heart of the case

Much of the filing focuses on how ChatGPT is designed. It points to human-like conversation, a tendency to agree with and flatter users — described in the complaint as "sycophancy" — personalisation and persistent memory as features engineered to increase engagement, noting that memory was switched on by default in 2024. It also cites OpenAI's own Model Spec, which the state says instructs the chatbot to avoid ending conversations.

Oregon contrasts those design choices with the company's public statements, including a May 2024 safety update in which OpenAI said it had built "strong default guardrails" into ChatGPT to protect children, and usage policies stating the service must never be used to endanger anyone under 18.

The lawsuit also targets ChatGPT for Teens, launched on August 18, 2026, and cites an October 7 assessment by Common Sense Media. According to that assessment, testers on more than a dozen newly created, parent-linked accounts spent up to an hour discussing suicidal thoughts, self-harm or disordered eating without a single alert reaching a parent. OpenAI has disputed the assessment's methodology.

The complaint further links ChatGPT to the deaths of several young people whose families are already suing the company, including Adam Raine, 16, and Amaurie Lacey, 17, and claims the company's systems flagged conversations by the attacker in the February 2026 school shooting in Tumbler Ridge, British Columbia, without leadership referring them to police. British Columbia has made a similar accusation in its own lawsuit.

Part of a much bigger legal wave

Oregon is the third state to sue OpenAI in five weeks: Texas filed on October 23 and California on November 16, each under a different mix of laws, and Florida brought the first state case in June. Private lawsuits are piling up in parallel — by one unofficial litigation tracker, 168 tort lawsuits had been filed against AI companies in the US by mid-November, with OpenAI facing more than any other developer.

OpenAI had not commented on the Oregon case at the time of reporting, according to Engadget. Responding to the state lawsuits filed in October, the company said it had "built industry-leading age-specific protections into our product that many other AI products lack," and that it supports safety policies applied across the whole AI industry rather than to a single developer.

The stakes go beyond one company. Oregon has a new chatbot law taking effect in January, and the case will test whether consumer protection and public nuisance law can impose design obligations on chatbot makers. For parents, the immediate takeaway is simpler: regulators on both sides of the Atlantic are increasingly sceptical of chatbot safety claims, and the parental controls on any AI app are worth checking yourself rather than taking on trust.

Sources: Oregon Department of Justice — "Keeping Kids Safe" initiative and the OpenAI legal action (November 2026), MLex — "OpenAI faces $2 billion Oregon suit over alleged harms to children" (November 17, 2026), Digital Watch Observatory — "Oregon sues OpenAI for $2bn over ChatGPT child safety" (November 2026), Bloomberg Law — reporting on the state lawsuits against OpenAI (October–November 2026), Common Sense Media — assessment of ChatGPT for Teens (October 7, 2026)

Advertisement
Advertisement