On October 23, 2026, Texas Attorney General Ken Paxton filed an original petition in Collin County against OpenAI Global, LLC, affiliated OpenAI entities, and Sam Altman. The petition and the same-day Office of the Attorney General release allege ChatGPT was marketed as safe and reliable without disclosing inaccurate or harmful outputs, and without the child safeguards the State says the DTPA and the SCOPE Act require.
Chatbot providers, developers and deployers that market safety, reliability or minor-protection claims to Texas users must review those claims immediately against the practices the petition alleges, or the Texas Office of the Attorney General can sue for an injunction and civil penalties of up to $10,000 per violation under section 17.47(c)(1).
- Authority: Texas Attorney General Ken Paxton, Consumer Protection Division
- Entity: OpenAI Global, LLC, affiliated OpenAI entities, and Sam Altman
- Measure and amount: civil petition for an injunction and up to $10,000 per violation; none assessed at filing
- Legal basis: Tex. Bus. & Com. Code sections 17.46(a), (b) and 17.47; SCOPE Act sections 509.051, 509.052, 509.054, 509.101 and 509.151
- Decision date: October 23, 2026 (petition and release the same day; cause number blank)
- Official source: State of Texas original petition (PDF)
What happened
The State sued under section 17.47. Defendants are six OpenAI entities and Sam Altman. Venue is Collin County. The cause number is blank. Relief sought is an injunction against offering ChatGPT to minors in Texas and against a new model release to Texas users before a court-approved evaluator reviews its guardrails, plus $10,000 per violation, disgorgement, costs and fees.
The breach
Count I cites sections 17.46(a) and (b). Paragraph 193 says Texas may seek an injunction and a civil penalty of up to $10,000 per violation. Listed practices: failure to warn of ChatGPT's dangers, content called unsuitable for children without age verification or parental consent, compulsive-use features, inadequate safety testing, and false claims of safety, reliability, or human emotions. Paragraph 205 cites section 17.47(c)(1). No separate amount is pleaded for consumers aged 65 or older.
Count II pleads the SCOPE Act, Chapter 509, sections 509.051, 509.052, 509.054 and 509.101: no commercially reasonable check that a user is a parent of a known minor, and no tools to limit sharing of that minor's identifying information or targeted advertising. Section 509.151 treats a SCOPE Act violation as a deceptive act under the DTPA, which the petition calls a per se breach of section 17.46(a).
Who else is in the same regime
OpenAI is the defendant, not the audience. Consumer-facing chatbot providers selling into Texas are in the same DTPA and SCOPE Act regime when they market safety, reliability or minor-protection claims. So are developers that market those claims to Texas users, and Texas-facing deployers that repeat a provider's safety claims in their own copy.
What to check now
- Compare safety, reliability and minor-protection claims shown to Texas users with the failure-to-warn practices in Count I.
- Check whether a Texas user under 18 can open an account without age verification or parental consent (sections 509.051 and 509.052).
- Confirm a verified parent can limit sharing of a known minor's identifying information and targeted ads (section 509.054), and how parent status is verified (section 509.101).
- If you deploy a third-party chatbot, do not repeat the provider's safety claims in your own copy without a separate review.
- Read injunction language as relief requested, not as an order.
