What the ChatGPT medical advice lawsuit alleges
The complaint alleges that ChatGPT-4o told a former pastor his groin tenderness was not dangerous, and that the pain turned out to be a pulmonary embolism that nearly killed him. Scott Winters filed it on July 22, 2026, in the Superior Court of California, County of San Francisco, against OpenAI and Sam Altman.
The Tech Justice Law press release says he nearly died after months of receiving inaccurate medical advice from ChatGPT-4o, "as described by the complaint." Everything in this section is what the complaint or the plaintiff's press release says, not something a court has found.
The complaint opens with the emergency. It says Winters called 911 and was taken to the nearest hospital, where he was in and out of consciousness and had violent convulsions.
He was rushed to the intensive care unit and placed under the care of the hospital's Critical Care Management team. The condition at issue is a pulmonary embolism, a blockage in an artery of the lung. The press release's title ties it to life-threatening blood clots.
Before that, the complaint says, Winters had been asking ChatGPT-4o about his symptoms and whether he should go to the hospital. He told it he felt odd twinges in different parts of his body and a tenderness in his groin.
The complaint says the model assured him the tenderness was "very likely another minor piece of the long story" of his gut and pelvic floor sensitivity, and "not something dangerous." It then says the pain in his groin was the onset of the embolism.
The complaint also describes other exchanges. It says ChatGPT-4o told Winters that his recurring dizzy spells and bouts of blood pressure instability did not yet meet the threshold for a serious condition.
It says the model suggested he stay home, "recliner-bound," and avoid too much movement until he had at least 8-10 more "episodes." According to the complaint, Winters adhered to that guidance, stayed in his recliner and limited his movement.
The complaint says his doctors stated the condition was likely brought on because of his immobility. The link between that immobility and ChatGPT-4o's guidance is the complaint's own.
The case at a glance
| Item | Detail |
|---|---|
| Plaintiff (the person suing) | Scott Winters, former pastor, Florida |
| Defendants | OpenAI Inc., OpenAI OpCo LLC, OpenAI Holdings LLC, Samuel Altman, plus unnamed employees and investors (Doe defendants) |
| Court | Superior Court of California, County of San Francisco |
| Filed | July 22, 2026 |
| Model at issue | ChatGPT-4o |
| Causes of action | 8, with a demand for a jury trial |
Why the complaint centers on Winters's faith
The complaint says the model used Winters's religious identity to keep him engaged with ChatGPT. It alleges that the model, drawing on his deep spiritual beliefs and trust in God, worked elements of his spirituality into its answers, including the line "God did not design your body to endlessly fail." It also alleges that, to optimize user engagement, the model sought to "ground" him in his faith by reminding him that he should not be particularly moved by seemingly "harmless sensations" and that God would carry him.
The press release adds that the complaint claims OpenAI exploited sensitive information about Winters's religious beliefs to increase engagement with the product. Those are claims by the plaintiff's side, and OpenAI's answer to them is not in the reports. Winters's own words appear in the same press release.
He says: "ChatGPT manipulated my own language and beliefs because it knew I was a pastor." He also says he nearly died and lost his job, his career, his ministry and his home.
"SUPERIOR COURT OF THE STATE OF CALIFORNIA FOR THE COUNTY OF SAN FRANCISCO"／"JURY DEMAND"／"John Doe Employees 1-10, and John Doe Investors 1-10"／"He frantically called 911 and was transported to the nearest hospital."／"As he was being wheeled through the hospital doors, he was in and out of consciousness and experiencing violent convulsions."／"He was rushed to the intensive care unit and placed under the care of the hospital’s Critical Care Management team."／"asking for input on his symptoms and advice about whether he should go to the hospital."／"Scott told ChatGPT-4o that he felt odd twinges in different parts of his body and that a tenderness"／"ChatGPT-4o assured him that this tenderness was “very likely another minor piece of the long story” of Scott’s gut and pelvic floor sensitivity, “not something dangerous” that necessitated medical attention."／"It turned out that the pain Scott was feeling in his groin was the onset of the pulmonary embolism that would nearly kill him."／"ChatGPT-4o told Scott that his recurring dizzy spells and bouts of blood pressure instability did not yet meet the threshold for being considered a serious condition."／"ChatGPT-4o suggested that it would be in Scott’s best interest to stay home, “recliner-bound” and avoid too much movement until he had at least 8-10 more “episodes.”"／"Scott adhered to ChatGPT-4o’s guidance to continue in his recliner-bound state and to limit movement."／"one that his doctors stated was likely brought on because of his immobility."／"drawing on Scott’s deep spiritual beliefs and trust in God"／"asserting that “God did not design your body to endlessly fail.”"／"To optimize user engagement, ChatGPT-4o sought to “ground” Scott in his faith by reminding him that he should not be particularly moved by seemingly “harmless sensations” and that God would carry"View official source →
"ChatGPT manipulated my own language and beliefs because it knew I was a pastor."／"Not only did I nearly die, but I also lost my job, my career, my ministry, my home, everything."／"exploited sensitive information about Winters’ religious beliefs to increase engagement with the product."View official source →
The eight causes of action in Winters v. OpenAI
The complaint brings eight causes of action, which are the separate legal grounds on which Winters asks the court to hold the defendants responsible. Two are strict liability claims, which hold a maker responsible for a defective product without requiring proof of carelessness.
Three are negligence claims, which say a party failed to take reasonable care. The other three are one claim under California's unfair competition law, one for negligent undertaking and one for invasion of privacy.
The table lists the counts in the order the complaint gives them. Each row repeats the complaint's own heading, with two additions from Courthouse News: the unfair competition label in row 6 and the note on Altman in row 7.
| # | Cause of action |
|---|---|
| 1 | Strict liability — defective design |
| 2 | Strict liability — failure to warn |
| 3 | Negligence — design defect |
| 4 | Negligence — failure to warn |
| 5 | Negligence — Cal. Bus. & Prof. Code §§ 2052, 4999.9 |
| 6 | Violation of Cal. Bus. & Prof. Code § 17200 et seq. (unfair competition law) |
| 7 | Negligent undertaking — against Sam Altman |
| 8 | Invasion of privacy under Cal. Const. Art. I, § 1 |
The fifth count is a negligence claim that cites sections 2052 and 4999.9 of the California Business and Professions Code. Separately, the press release says the complaint claims OpenAI engaged in the unauthorized practice of medicine. The seventh count is the one aimed at Altman personally: Courthouse News reports that Winters claims negligent undertaking against him, a claim that someone who took on the job of keeping others safe did it carelessly.
The press release describes the eight counts as including negligence claims and violations of California consumer protection and privacy laws. The eighth count rests on the privacy right in the California Constitution. These descriptions come from the complaint and from the plaintiff's own press release.
"FIRST CAUSE OF ACTION STRICT LIABILITY (DEFECTIVE DESIGN)"／"SECOND CAUSE OF ACTION STRICT LIABILITY (FAILURE TO WARN)"／"THIRD CAUSE OF ACTION NEGLIGENCE (DESIGN DEFECT)"／"FOURTH CAUSE OF ACTION NEGLIGENCE (FAILURE TO WARN)"／"FIFTH CAUSE OF ACTION NEGLIGENCE (CAL. BUS. & PROF. CODE §§ 2052, 4999.9)"／"SIXTH CAUSE OF ACTION VIOLATION OF CAL. BUS. & PROF. CODE § 17200 et seq."／"SEVENTH CAUSE OF ACTION NEGLIGENT UNDERTAKING"／"EIGHTH CAUSE OF ACTION INVASION OF PRIVACY UNDER CALIFORNIA CONST."View official source →
"The complaint further claims OpenAI engaged in the unauthorized practice of medicine and exploited sensitive information about Winters’ religious beliefs to increase engagement with the product."／"Winters’ lawsuit brings eight causes of action against OpenAI and the other defendants, including negligence claims and violations of California consumer protection and privacy laws."View official source →
"Winters’ claims include strict liability, negligence, violation of California’s unfair competition law and invasion of privacy."／"He also claims negligent undertaking against Altman."View official source →
What Winters asks the court to order on GPT-4o and ChatGPT Health
Winters asks for damages and for court orders that would change how GPT-4o and ChatGPT Health operate, and these are requests, not rulings. An injunction is a court order telling a company to do something or stop doing it. Courthouse News reports that the complaint asks for the following:
- automatic conversation termination for self-harm content or when immediate medical assistance is necessary
- safety warnings that ChatGPT-4o may lead to physical dependencies and generate harmful medical advice
- removal of the GPT-4o model and deletion of training data built from GPT-4o
- information for users about the data sources used to train generative AI models
- a pause on ChatGPT Health until independent safety audits find it safe
The first item would change what ChatGPT does when a user may need urgent care, and the second would add warnings about dependence and harmful medical advice. The next two go to the model and the data behind it. The last targets a separate product: Courthouse News says OpenAI announced ChatGPT Health in January as a dedicated experience in ChatGPT designed for health and wellness.
"He seeks damages and an injunction ordering OpenAI to implement automatic conversation termination for self-harm content or when immediate medical assistance is necessary and safety warnings that ChatGPT-4o may lead to physical dependencies and generate harmful medical advice."／"He also seeks an order directing OpenAI to get rid of the GPT-4o model, delete training data built from GPT-4o and provide users with information about data sources used to train generative AI models."／"Winters wants the court to pause the operation of ChatGPT Health"／"which OpenAI announced in January as"／"a dedicated experience in ChatGPT designed for health and wellness"／"independent third parties determine the product to be safe through comprehensive safety audits."View official source →
What OpenAI has said and what is still unproven
Nothing in the complaint has been proven, and OpenAI did not immediately respond to a request for comment on July 22, 2026. Courthouse News reported that representatives for OpenAI did not immediately respond to a request for comment.
A complaint is one side's account, and everything above about what happened to Winters comes from it or from the plaintiff's press release. The reports record no court ruling on any claim.
The press release also carries statements from Winters's counsel. Matthew P. Bergman of the Social Media Victims Law Center says: "If ChatGPT were a physician dispensing medical advice, it would be guilty of medical malpractice."
He also says OpenAI actively promotes itself as a reliable source of medical guidance. These are counsel's arguments for the plaintiff, not findings.
"Representatives for OpenAI did not immediately respond to a request for comment."View official source →
"Scott Winters nearly died because ChatGPT acted like a medical authority while having none of the responsibility,"／"If ChatGPT were a physician dispensing medical advice, it would be guilty of medical malpractice."／"OpenAI actively promotes itself as a reliable source of medical guidance and must not be allowed to hide behind its algorithm to avoid accountability for providing false and misleading medical information to users."View official source →
How this suit fits with other lawsuits against OpenAI
Winters v. OpenAI is one of several lawsuits over harms attributed to ChatGPT, according to Courthouse News. The outlet reported that several other lawsuits against OpenAI have been filed in San Francisco Superior Court and across the country.
Tech Justice Law says in the press release that it has previously brought similar lawsuits against OpenAI, alleging a pattern of inadequate guardrails to protect sensitive users. It also describes itself as a co-filer of the first-ever lawsuits against a chatbot product that Character AI developed with support from Google.
Copyright suits against other AI companies are covered in "AI Copyright Lawsuits: Perplexity, Reddit, Brave," which gives the wider litigation background. That article concerns different defendants and different claims from this case. The primary documents for this case are the complaint and the Tech Justice Law press release, both linked in the source boxes above.
"Tech Justice Law has previously brought similar lawsuits against OpenAI, alleging a pattern of inadequate guardrails to protect sensitive users."／"TJL co-filed the first-ever, groundbreaking lawsuits against a popular, “AI” chatbot product developed by Character AI, with support by Google, raising public awareness of chatbots’ real-world harms."View official source →
"Several other lawsuits against OpenAI have been filed in San Francisco Superior Court and across the country over purported harms derived from ChatGPT"View official source →
FAQ
- Who filed the ChatGPT medical advice lawsuit, and against whom?
- Scott Winters, a former pastor and a Florida resident, filed suit in San Francisco County Superior Court on July 22, 2026. The complaint names three OpenAI entities, Sam Altman personally, and unnamed employees and investors as defendants. His counsel are Tech Justice Law, the Social Media Victims Law Center and the Institute for Law, Innovation & Technology at Temple University.
- What does the complaint say ChatGPT actually told him?
- The complaint says that when Winters mentioned tenderness in his groin, ChatGPT-4o assured him it was very likely another minor piece of a long story about his gut and pelvic floor sensitivity, and not something dangerous that needed medical attention. It also quotes the model invoking his faith. These are the complaint's allegations about the model's outputs.
- How many causes of action does the complaint bring?
- Eight. Two are in strict liability, for defective design and failure to warn. Three are in negligence, for design defect, failure to warn, and a claim citing California Business and Professions Code sections 2052 and 4999.9. The remaining three are a claim under the state's unfair competition law, a negligent undertaking claim against Altman personally, and an invasion of privacy claim under the state constitution.
- What remedies is he asking for?
- He asks for damages and an injunction, which is a court order telling OpenAI to act or stop acting. The requested orders are automatic conversation termination when someone needs immediate medical help, warnings about dependence and harmful medical advice, removal of GPT-4o and the training data built from it, and information about training data sources. He also asks the court to pause ChatGPT Health. These are requests, not orders the court has granted.
- Has OpenAI responded?
- Courthouse News Service reported on July 22, 2026, that representatives for OpenAI did not immediately respond to a request for comment. The reports and filings behind this article include no later statement from OpenAI. Everything in the complaint is the plaintiff's allegation, and none of it has been tested in court.
Related articles
OpenAI, ServiceNow expand deal with ChatGPT pluginOpenAI and ServiceNow expanded their deal on Oct 28, 2026. A ServiceNow MCP plugin for ChatGPT is live now; Codex, security and voice tools follow in 2027.
ChatGPT and Codex usage limits merge in DecemberFrom December, Plus, Pro and Business plans will share one token-based allowance across ChatGPT and Codex. Instant chat won't count, and no start date is set.
ChatGPT now works with iCloud Mail and CalendarOpenAI added iCloud Mail and Calendar plugins to ChatGPT on Oct 27 for all plans. You connect with an Apple app-specific password that outlives a disconnect.
ChatGPT limits weapons information for teensOpenAI said on Oct 30, 2026 that ChatGPT will limit weapons information for teens, end chats fixated on violence, and report on safety every six months.