Texas has sued OpenAI and its chief executive, Sam Altman, and the first thing a judge will read is not a statute. It is Altman, in italics: “AI will probably most likely lead to the end of the world, but in the meantime, there’ll be great companies.” The state puts the line at the top of its petition as evidence, not decoration. The argument that follows is simple: a company whose founder talks openly about the technology’s dangers cannot also sell ChatGPT to families as a safe tool.

That is the mechanism worth watching. Most coverage has run the short version, that Texas wants to keep kids off ChatGPT. The petition reaches further. It asks a court to decide who may use the chatbot, what it may say, and whether OpenAI’s next model may be released in the state at all.

What Texas wants a court to order

Attorney General Ken Paxton filed the case in Collin County district court under the Texas Deceptive Trade Practices Act and the SCOPE Act, the state’s children’s online safety law. It names six OpenAI entities and Altman himself. The state wants him held personally liable, alleging that as CEO he “has personally chosen which ChatGPT features are greenlit and put into production.” The Record called that an unusual move.

The list of requested injunctions is long. One line asks the court to stop OpenAI “offering ChatGPT to minors in Texas”, and as written it carries no exception for parental consent or school use. Others would require age verification and block the chatbot from serving tips on eating disorders, self-harm and mass murder. Then comes the clause with the widest reach, which would bar OpenAI from “releasing any new artificial intelligence model to Texas users unless an independent third-party evaluator, approved by the Court, has reviewed its safety guardrails and approved its release.”

A release key held by a court

The minors ban gets the headlines; the pre-release clause is the bigger idea. When eight other states sued OpenAI earlier this month, Kansas Attorney General Kris Kobach said Congress had shown little appetite for restraining AI companies, which leaves them to police themselves. “We are not going to hold our breath,” he said. Texas is testing a different answer: a court-approved evaluator holding the launch button, one state at a time.

There is also a tension inside the filing. The SCOPE Act it invokes does not ban minors from digital services. According to the petition’s own summary, it requires that verified parents be able to control a known minor’s privacy and account settings, and to monitor and limit their time on the service. That is a rulebook for supervised use. The remedy Texas asks for is stricter than the statute’s model, and it echoes Common Sense Media, which the petition quotes: until OpenAI proves its teen safeguards with independent testing, “ChatGPT should be for adults only.”

What the petition says goes wrong for teenagers

The factual case leans on age checks and parental tools. According to the petition, as summarized by The Next Web, the free version of ChatGPT has no age verification, the paid version asks for an age without checking it, and the parental controls OpenAI added in September 2025 are optional. Common Sense Media found that ChatGPT for Teens, launched in August, “could give parents false confidence in guardrails and safety alerts that frequently don’t work.” The filing also cites the death of Adam Raine, a 16-year-old whose parents sued OpenAI and Altman in California.

These are allegations. No court has ruled, and nothing changes for ChatGPT users in Texas unless a judge grants the injunctions. OpenAI’s reply matches what it told the other states: it has “worked closely with experts to continually improve ChatGPT, mitigate risks and build additional safeguards,” and wants “pragmatic AI policies that apply to the entire AI industry – not just one company.” It points to age-prediction tools, break reminders and a separate teen experience with parental oversight.

Nine states, nine sets of rules

Texas filed on October 23, the ninth state to take OpenAI to court this month. Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia filed separate consumer-protection suits on October 14, accusing the company of tuning ChatGPT toward agreement and engagement while misleading parents about its safety. Each case runs under its own state law. Read that way, OpenAI’s “not just one company” line asks for one industry rulebook instead of a stack of state court orders.

Texas also wants civil penalties of $10,000 per violation, the disgorgement of OpenAI’s assets and a jury trial. Its petition notes that OpenAI went from a valuation of about $17 billion to more than $1.4 trillion in four years, and calls that rise “a web of deceit.” Paxton’s office has squeezed big tech before: Meta agreed in July 2024 to pay Texas $1.4 billion to settle claims that it used facial recognition without users’ consent.

The quote at the top of the petition dates from 2015, when AI was, in the filing’s own words, “a virtually unknown concept outside of science fiction.” Texas now wants a jury to read it as a warning OpenAI gave itself and, the state says, never passed on to the people typing into ChatGPT.

Tags: OpenAI, Sam Altman, chatgpt, AI regulation, texas