Oregon OpenAI Lawsuit Seeks $2.035 Billion, Pleading Negligence and Nuisance on Top of ChatGPT Penalty Claims
The Oregon OpenAI lawsuit, filed November 17, 2026 in Multnomah County Circuit Court, seeks $2.035 billion and an injunction over ChatGPT's safety for minors, and adds negligence and public nuisance damages claims that the October multistate suits largely left out.
Torts Desk··13 min read

The Oregon OpenAI lawsuit, filed November 17, 2026 by Attorney General Dan Rayfield in Multnomah County Circuit Court, accuses OpenAI of endangering children and misleading parents about ChatGPT's safety and asks for at least $2.035 billion plus a court order forcing real safeguards for minors. It makes Oregon the tenth state to sue the company this year, and the first since the October 14 multistate wave to put a single, headline damages figure on the harm, built on negligence and public nuisance counts that sit alongside the usual consumer protection claim.
That pairing matters more than the size of the number. Most state actions against OpenAI so far are penalty cases: they multiply a statutory per-violation cap by whatever count a court eventually accepts. Oregon's complaint keeps that penalty track under the Unlawful Trade Practices Act but adds tort theories that measure money by injury to the public, which is closer to how private plaintiffs and the opioid and social media nuisance cases frame recovery. For anyone tracking chatbot injury litigation as an emerging Mass Tort, the Oregon OpenAI lawsuit is the clearest sign yet that state enforcers are borrowing the product liability playbook.
Why is Oregon suing OpenAI?
Oregon alleges that OpenAI knew ChatGPT lacked the guardrails needed to protect minors, kept marketing it to them anyway, and misrepresented its safety to parents. Rayfield summarized the case in one line: "OpenAI knows that ChatGPT is harming children, and has failed to meaningfully fix their dangerous platform."
The allegations, as described in the Oregon Department of Justice release and in reporting by OPB, the Oregon Capital Chronicle, KLCC and Bloomberg Law, fall into four groups:
- Design that fosters dependence. The state says OpenAI knowingly and deliberately engineered ChatGPT to foster dependence among users, using human-like conversation and "sycophantic" answers that earn trust and keep young users engaged. Coverage of the complaint describes the growth push as making the chatbot overly affirming, which the state says is most dangerous when a child raises a sensitive subject.
- Marketing to minors. Oregon alleges OpenAI pursued young users directly, including through an experience called ChatGPT for Teens, using what the complaint calls unfair tactics.
- Weak age and consent controls. OpenAI's terms allow users as young as 13 with a parent's or guardian's consent, but the complaint says the company often fails to confirm that a parent actually signed off.
- Undisclosed risk. The state says OpenAI sold a product with safety risks it did not disclose, while representing the product as safe and age-appropriate.
Rayfield told reporters that no specific incident in Oregon prompted the suit. That is a deliberate posture: the case is framed as statewide consumer and public harm rather than as an attempt to litigate one family's injury, which keeps the state's claims distinct from the private wrongful death actions pending in California.
How much is Oregon seeking from OpenAI?
Oregon is seeking at least $2.035 billion in damages, plus civil penalties for each willful violation of the Unlawful Trade Practices Act, an injunction and other equitable relief. Some outlets rounded the figure to "$2 billion"; the state's own release and most Oregon outlets use $2.035 billion.
The relief requested breaks down this way, based on the release and consistent reporting:
| Relief | Legal basis | What the state asks for |
|---|---|---|
| Damages of at least $2.035 billion | Negligence; public nuisance; unjust enrichment | Compensation for Oregonians and abatement of the alleged harm to youth |
| Civil penalties | UTPA, ORS 646.608 and 646.642 | Up to $25,000 per willful violation, amount set by the court |
| Permanent injunction | UTPA, ORS 646.632; court's equitable power | Bar OpenAI from misrepresenting the safety and age-appropriateness of ChatGPT and ChatGPT for Teens |
| Safety measures | Injunction and nuisance abatement | Effective age assurance and safeguards against harmful content for minors |
| Disgorgement of unfair profits | Unjust enrichment | Return of revenue the state says was earned at Oregonians' expense |
One summary of the complaint reports that it names six OpenAI entities as defendants and demands a jury trial, and that the $2.035 billion floor attaches to the negligence and public nuisance counts. That detail was not confirmed by a second source at publication.
How did Oregon calculate the $2 billion figure?
Rayfield told KATU that his office arrived at $2.035 billion based on the number of Oregon children who use or have access to ChatGPT. Neither the release nor any report published the child count or the per-child multiplier, so the exact formula is not public.
Two observations follow from what is public. First, a population-based figure is a damages theory, not a penalty theory: it values the harm to a class of residents rather than counting discrete deceptive acts. Second, it can be cross-checked against the penalty track. At the UTPA ceiling of $25,000 per willful violation, $2.035 billion would equal 81,400 violations, which shows how quickly a per-violation count could approach the damages figure if a court counted violations per minor user. How Oregon courts define a single "violation" for a product used by hundreds of thousands of residents is an open question the case will likely have to answer.
What laws does Oregon say OpenAI violated?
Oregon pleads one statute and three common-law theories:
- Unlawful Trade Practices Act (UTPA). The core claim is that OpenAI misrepresented ChatGPT's safety for minors, marketed it to youth using unfair tactics, and sold a product with undisclosed safety risks. The UTPA lists unlawful practices in ORS 646.608 and lets the attorney general seek injunctions under ORS 646.632. Under ORS 646.642, where a person willfully uses an unlawful practice, the court may impose a civil penalty "not exceeding $25,000 per violation." The complaint alleges the violations were willful, which is the trigger for that penalty.
- Negligence. The state says OpenAI breached a duty of care owed to minors through its design and deployment choices. This is the count closest to the Design Defect and Failure to Warn theories in private chatbot cases.
- Unjust enrichment. Described in the release as unfair profit at Oregonians' expense, this supports disgorgement.
- Public nuisance. The state alleges ChatGPT's deployment to minors created a public nuisance threatening youth well-being. Public nuisance was the backbone of the state opioid litigation and is now a staple of state social media suits, though its reach over products, rather than land use, remains contested from state to state.
Compared with the October 14 filings, Oregon's mix is closest to Arkansas, which also pleaded negligence, design defect, failure to warn, unjust enrichment and public nuisance alongside its deceptive trade practices claim. Most of the other October states sued under consumer protection statutes alone.
What is ChatGPT for Teens and why does Oregon target it?
ChatGPT for Teens is OpenAI's age-specific experience for younger users, which the company presents as having stricter protections. Oregon's complaint treats it as evidence of direct marketing to minors rather than as a mitigation, and the requested injunction names it expressly: the state wants OpenAI barred from misrepresenting the safety and age-appropriateness of both ChatGPT and ChatGPT for Teens.
One summary of the complaint reports that the teen experience launched in August 2026, that linking a parent account to it is optional, and that before August 2026 ChatGPT had no parental consent verification at sign-up. Those details come from a single secondary account and had not been independently confirmed at publication. The theory they support is consistent with the confirmed allegations, though: a safety-branded product that can be bypassed is, in the state's framing, a misrepresentation rather than a defense.
How many states have sued OpenAI over ChatGPT?
Ten states have now sued OpenAI over ChatGPT in 2026, all in their own state courts.
| State | Filed | Forum | Notable feature |
|---|---|---|---|
| Florida | June 1, 2026 | Tenth Judicial Circuit (Highlands County) | FDUTPA and negligence theories; names Sam Altman; temporary injunction motion filed Sept. 28, 2026 |
| Arizona | Oct. 14, 2026 | Maricopa County Superior Court | Consumer Fraud Act; up to $10,000 per willful violation |
| Arkansas | Oct. 14, 2026 | Faulkner County Circuit Court | Deceptive trade practices plus negligence, design defect, public nuisance; jury demand |
| Idaho | Oct. 14, 2026 | Ada County (Fourth Judicial District) | $5,000 per violation; restitution |
| Iowa | Oct. 14, 2026 | Polk County District Court | Up to $40,000 per violation |
| Kansas | Oct. 14, 2026 | Shawnee County (reported) | Kansas Consumer Protection Act |
| Montana | Oct. 14, 2026 | Lewis and Clark County | Consumer protection and data privacy |
| Nebraska | Oct. 14, 2026 | State district court | $50,000 per design-code violation |
| West Virginia | Oct. 14, 2026 | Mason County Circuit Court | Consumer protection and public nuisance; pleads it is not bound by user arbitration agreements |
| Oregon | Nov. 17, 2026 | Multnomah County Circuit Court | UTPA plus negligence, unjust enrichment, public nuisance; at least $2.035 billion |
Oregon did not join the October coordinated filing. Its separate timing and its decision to file in Multnomah County, Oregon's most populous county and home to Portland, give it its own judge, schedule and jury pool. Oregon's attorney general has acted on AI chatbots before, raising child safety and privacy objections in the Character.AI bankruptcy and joining other attorneys general in urging Congress to set AI safety rules.
Can OpenAI force Oregon into arbitration or move the case to federal court?
Almost certainly not into arbitration, and removal faces steep obstacles. ChatGPT's consumer terms send individual users' disputes to individual arbitration with a Class Action Waiver, but the State of Oregon never accepted those terms. Under EEOC v. Waffle House, Inc., 534 U.S. 279 (2002), a government plaintiff suing in its own right is not bound by an Arbitration Agreement it did not sign, and West Virginia's October complaint pleaded exactly that point. Civil penalties, injunctions and nuisance abatement belong to the state alone, so no user's contract can redirect them. The only plausible fight is at the margins, over whether any part of the damages ask is, in substance, individual users' claims.
Federal court is also hard to reach. A state is not a citizen for diversity purposes, so there is no diversity jurisdiction over a suit brought by Oregon in its own name. The Class Action Fairness Act (CAFA) does not help: in Mississippi ex rel. Hood v. AU Optronics Corp., 571 U.S. 161 (2014), the Supreme Court held that a state's parens patriae suit is not a CAFA "mass action" even when it seeks money for residents. And because every reported count arises under Oregon statutes or common law, defenses based on the First Amendment or Section 230 do not create federal-question jurisdiction under the well-pleaded complaint rule. The practical result is a tenth separate state-court docket, with no Judicial Panel on Multidistrict Litigation (JPML) available to centralize the state cases.
Does the Oregon lawsuit help families with private ChatGPT injury claims?
Indirectly. The Oregon case does not create a claims process, a settlement fund or a right for individual families to recover, and any damages awarded would go to or through the state. But it adds pressure on the same liability theories private plaintiffs are pursuing.
The private cases are concentrated in In re ChatGPT Product Liability Cases, JCCP No. 5431, a California coordinated proceeding in San Francisco Superior Court that includes Raine v. OpenAI, the 2025 wrongful death suit over a 16-year-old's suicide. Among the November 2025 suits filed by the Social Media Victims Law Center and Tech Justice Law Project was a wrongful death claim on behalf of Joe Ceccanti, 48, of Oregon. There is still no federal Multidistrict Litigation (MDL) for ChatGPT claims.
Three interactions are worth watching:
- Discovery. State attorneys general can obtain internal safety records through investigative demands and litigation discovery. If documents about sycophancy testing, age assurance or teen-mode design become public in Oregon, they can inform private Design Defect and Failure to Warn claims, subject to protective orders.
- Standard of care. An injunction or consent judgment requiring age assurance or crisis-escalation protocols does not bind OpenAI in private cases through preclusion, but it shapes what juries hear a reasonable company should have done.
- Nuisance as a template. If an Oregon court lets the public nuisance count survive a motion to dismiss, that ruling becomes persuasive authority for other states and for private plaintiffs testing nuisance theories against AI products.
What did OpenAI say about the Oregon lawsuit?
OpenAI spokesperson Drew Pusateri issued a statement nearly identical to the company's response to the October 14 suits. That earlier statement said: "While no system can be perfect, we'll continue to work with experts to inform our work and build additional safeguards, such as the parental controls, break reminders and age-appropriate model protections that exist today." The company has also described its age-specific protections as industry leading. No answer, motion to dismiss or removal notice had been reported in Multnomah County as of November 18, 2026.
What happens next in the Oregon OpenAI lawsuit?
The next steps follow the ordinary Oregon civil timetable. OpenAI must be served and will then respond, most likely with a motion to dismiss directed at the common-law counts. Expect three early fights:
- Duty and nuisance. Whether Oregon law recognizes a duty owed by a software provider to minors in this setting, and whether public nuisance extends to a digital product, are the most vulnerable parts of the complaint.
- Section 230 and the First Amendment. OpenAI may argue that chatbot output is protected speech or that some claims treat it as a publisher. Because ChatGPT generates its own output rather than hosting third-party content, Section 230 is a weaker defense here than in social media cases, but the argument will likely be raised.
- Counting violations. Any penalty phase will turn on how an Oregon court counts UTPA violations, the same question that produced a roughly 43.9 million-violation finding against Meta in New Mexico in September 2026.
The UTPA claims are likely to be the most durable, because they rest on representations about safety rather than on novel tort duties. The damages floor depends on the tort counts surviving.
What it means for practitioners
For personal injury and mass tort firms. The Oregon OpenAI lawsuit reinforces the liability narrative at the center of chatbot injury intake: engagement-maximizing design, inadequate age and consent controls, and safety marketing that outran the product. Firms with Oregon clients should track the Multnomah County docket for document productions, and should note that the state's case does not toll or resolve any individual's Statute of Limitations.
For defendants and AI developers. State-as-plaintiff litigation now spans ten forums and two remedial models, per-violation penalties and population-based damages. Compliance teams should treat the injunctive asks across all ten states (verified parental consent, effective age assurance, limits on sycophantic design, truthful safety claims for teen products) as a de facto regulatory baseline.
For arbitration clause drafters. No clause reaches a sovereign plaintiff, so the drafting question is narrower: how a consumer clause handles an individual user's claim for the same harm after a state has recovered on that user's behalf, and whether a Mass Arbitration Protocol anticipates follow-on demands timed to disclosures in state litigation.
For insurers and funders. A damages theory keyed to a state's child population is easier to scale than a count of deceptive statements. If other states copy Oregon's model, reserve and underwriting assumptions built on per-violation penalty caps will understate exposure.
FAQ
Why is Oregon suing OpenAI?
Oregon alleges OpenAI knowingly marketed ChatGPT to children while it lacked adequate safeguards, designed it to foster dependence through sycophantic responses, failed to verify parental consent for users aged 13 to 17, and misrepresented its safety, in violation of the Unlawful Trade Practices Act and Oregon common law.
How much is Oregon seeking from OpenAI?
At least $2.035 billion in damages, plus civil penalties of up to $25,000 per willful UTPA violation, disgorgement and an injunction requiring effective age assurance and truthful safety claims.
When and where was the Oregon OpenAI lawsuit filed?
Attorney General Dan Rayfield filed it on November 17, 2026 in Multnomah County Circuit Court, as the first action of Oregon's Keeping Kids Safe Initiative.
How many states have sued OpenAI?
Ten: Florida (June 1, 2026), Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia (October 14, 2026), and Oregon (November 17, 2026).
Can families join the Oregon lawsuit against OpenAI?
No. The state sues in its own name and individuals cannot join it as plaintiffs. Families with injury claims proceed separately, most of them in the coordinated California proceeding, JCCP No. 5431.
Has a court ruled on Oregon's claims?
No. The complaint was filed November 17, 2026, OpenAI disputes the allegations, and no court has made any finding.
Published for legal professionals. Analysis and summaries only — not legal advice, and no attorney-client relationship is created by use of this site.
Read next

States Sue OpenAI: Eight AGs Take ChatGPT to State Court, Where Its Arbitration Clause Can't Follow
Eight states sued OpenAI on Oct. 14, 2026 in eight separate state courts, alleging ChatGPT was sold as safe while harming children. The suits seek per-violation penalties up to $50,000, age verification and limits on AI testing, and no user arbitration clause binds them.
Policy Desk · 14 min

The ChatGPT Product Liability JCCP Reconvenes With 23 Cases, No MDL and No Arbitration Fight
In re ChatGPT Product Liability Cases, JCCP 5431, holds a further case management conference in San Francisco today, 23 September 2026, its first since leadership was appointed on 4 August. Twenty-three coordinated actions, no federal MDL, and no motion to compel arbitration anywhere on the docket.
Torts Desk · 15 min

Character.AI Bankruptcy: What Chapter 11 Means for the Chatbot Injury Lawsuits
Character.AI filed for Chapter 11 in Delaware on October 12, 2026, with a $20 million Disney stalking-horse bid and 40 attorneys general already in court. The automatic stay freezes the private chatbot injury suits against it, but not the claims against Google.
Torts Desk · 16 min