AI Chatbot Lawsuits: ChatGPT, Character.AI and Gemini Harm Claims

Families across the country have filed AI chatbot lawsuits against the companies behind the technology. The suits claim the chatbots were built to hold users’ attention and did not respond safely when those users were in crisis. The cases involve teenagers and adults who died by suicide, harmed themselves, or suffered severe psychiatric crises. The chatbots at the center of these suits include Character.AI, ChatGPT, and Google’s Gemini. If something like that happened in your family, we can review a potential claim at no cost. The Law Offices of Rudolph F.X. Migliore, P.C. works with a nationally recognized co-counsel network on these claims.

What the AI Chatbot Lawsuits Claim

Companion-style chatbots are designed to feel personal. They remember earlier conversations, use human mannerisms, and answer with warmth and agreement. According to the lawsuits, those choices drew vulnerable users, many of them teenagers, into heavy daily use and emotional dependence. Meanwhile, the products allegedly lacked basic safeguards for someone in crisis.

The specific allegations differ from case to case, but several themes keep coming up:

• A chatbot that agreed with and encouraged whatever the user said, including suicidal thoughts and paranoid or grandiose beliefs, instead of pushing back
• Conversations about self-harm that went on without the chatbot ending them or steering the person toward real help
• Characters and personas built to feel like a friend or romantic partner, which plaintiffs say deepened isolation and dependence
• No reliable way to confirm a user’s age or to report a dangerous conversation
• Products released, the lawsuits say, before their mental-health risks were adequately tested

Common legal claims include product liability (design defect and failure to warn), negligence, and wrongful death, often paired with state consumer-protection claims. Some news coverage calls the delusion cases “AI psychosis,” an informal label rather than a medical diagnosis. The companies dispute these claims, and they remain allegations unless proven in court.

Where the Cases Stand

Character.AI. One of the first Character.AI lawsuits, Garcia v. Character Technologies, was filed in federal court in Florida in October 2024. It was brought by the family of a teenage boy who died by suicide. In May 2025, the judge allowed most of the claims to proceed. The court ruled that the app could be treated as a product for claims aimed at its design, rather than at the ideas or words it produced. It also declined, at that early stage, to hold that the chatbot’s output was speech protected by the First Amendment. That decision came from a single trial court at the pleading stage, before any evidence was tested, and it does not bind other courts.

In November 2025, Character.AI stopped letting users under 18 hold open-ended chats with its characters. In January 2026, court filings showed that the company, its co-founders, and Google had agreed in principle to settle the Garcia case and other family lawsuits. Those included cases filed in Colorado, New York, and Texas. The terms were not made public. New suits have been filed against Character.AI since January, including one in federal court in California in August 2026.

ChatGPT (OpenAI). One of the earliest ChatGPT cases was filed in San Francisco in August 2025. It was brought by the parents of a California teenager who died by suicide. In its response, OpenAI denied that ChatGPT caused his death. It argued that he had misused the product in violation of its terms of use. Seven more suits followed in November 2025, most of them on behalf of adults. Four involve deaths by suicide, and the rest claim the chatbot fed delusions that led to serious harm. Many of these cases focus on GPT-4o, a ChatGPT model the plaintiffs say was rushed out despite internal warnings about its risks.

In February 2026, a California court coordinated the state-court cases into a single proceeding in San Francisco, the ChatGPT Product Liability Cases (JCCP No. 5431). A coordinated proceeding is California’s way of managing related lawsuits together, similar in purpose to a federal multidistrict litigation. Other ChatGPT cases are pending in federal court.

Gemini (Google). In March 2026, a father whose adult son died by suicide sued Google in federal court in California. He claims Google’s Gemini chatbot fed his son’s delusions. Google has defended Gemini’s safeguards and says the chatbot repeatedly referred his son to a crisis hotline.

Regulators. Government agencies are moving in parallel. In September 2025, the Federal Trade Commission ordered seven companies to report how they measure and monitor their chatbots’ effects on children and teens. The companies included OpenAI, Character Technologies (the company behind Character.AI), Alphabet, and Meta. Kentucky’s Attorney General sued Character Technologies in January 2026, and Florida’s Attorney General sued OpenAI and its CEO in June 2026. These government actions are separate from individual injury claims and do not resolve them.

No federal multidistrict litigation (MDL) has been created for AI chatbot suicide, self-harm, or psychiatric-injury claims as of October 2026. Cases are moving forward one at a time in federal courts and, for ChatGPT, through the coordinated California proceeding. This area is changing quickly, so contact us for the current status.

Every case is different. These lawsuits generally allege that a chatbot’s design contributed to a serious mental-health injury. As the CDC points out, suicide rarely traces back to a single cause. Proving causation means looking at the whole picture: the person’s history, medical and counseling records, the chatbot conversations, and other evidence specific to the case.

The New York Connection

New York was one of the first states to put safety rules on AI companions. Article 47 of the state’s General Business Law took effect on November 5, 2025. It applies to chatbots that meet the law’s definition of an “AI companion,” which excludes certain products designed mainly for efficiency, research, technical assistance, or customer service. A covered company must have a protocol to make reasonable efforts to detect and address expressions of suicidal thinking or self-harm. When the chatbot detects them, it must refer the user to crisis services such as the 988 Lifeline. Users must also receive clear, periodic notices that they are not talking to a human. The New York Attorney General enforces the law; it does not expressly give individuals a right to sue. The injury lawsuits described on this page rest on product-liability, negligence, and similar claims instead.

New Yorkers are part of the litigation, too. A New York family’s federal lawsuit against Character.AI, filed in Albany in September 2025, reached a settlement in principle in January 2026. Our firm evaluates these claims for Long Island and New York families and works with our co-counsel network when a case moves forward. How an individual case proceeds, including which state’s law applies, is something we evaluate at intake.

Frequently Asked Questions: AI Chatbot Claims

Talk With Us About an AI Chatbot Claim

If chatbot use was followed by a suicide, a suicide attempt or self-harm, or a psychiatric crisis in your family, please reach out. The Law Offices of Rudolph F.X. Migliore, P.C. evaluates these claims at no cost, from Long Island, for families throughout New York. Call 631-543-3663 for a free case evaluation, or use the contact form below. There is no fee unless we recover compensation for you.

This page is for general information and is not legal advice. The lawsuits described here contain allegations that have not been proven in court, and settlements are not admissions of liability. It reflects developments as of October 2026, and this litigation is changing quickly. Each case depends on its own facts, and no outcome can be guaranteed.