A Lebanese family sues OpenAI over their son's death: what the Habib case alleges

Content note: this article discusses suicide. If you or someone you know is struggling, help is listed at the end.

On 22 April 2026, Mathias Habib, a 15-year-old boy from Lebanon, took his own life. Six months later his parents, Charbel Habib and Soha Samaha, filed a wrongful death lawsuit against OpenAI, its affiliated companies and its chief executive, Sam Altman, in the Superior Court of California, County of San Francisco. The family says ChatGPT, the chatbot their son had used for three years, played a direct part in his death.

The allegations below come from the complaint and from the family's interviews with Al Arabiya and LBCI. No court has ruled on them. At the time of writing, we could not find a public response from OpenAI to this case.

A boy who did not fit the usual picture

Charbel Habib described his son to Al Arabiya as a strong student, well liked by his classmates, with no history of mental health problems. Mathias hiked and played the guitar, and his father says he did not like spending long hours on his phone. The day before he died, the family had lunch together and talked about an upcoming trip.

According to the family, Mathias began using ChatGPT about three years before his death. He used it the way many teenagers do: homework questions, books he was reading, cameras, hiking routes, classic cars. His parents saw nothing in that history that suggested heavy or late-night use.

How the conversations changed

The family's account, as set out in the lawsuit, centres on the last three months of his life. In January 2026, Mathias started reading Marcus Aurelius's Meditations and asked ChatGPT whether Stoicism was a cult. Over the following weeks, the family alleges, the conversations moved from Stoicism to Nietzsche, Camus, Schopenhauer, Epicurus and nihilism. His father claims the chatbot raised several of these topics itself, and he describes this as an engagement tactic designed to keep the conversation going.

The complaint quotes ChatGPT telling Mathias that "the unexamined life is not worth living", and describing suicide as "calm and dignified". The family stresses that ChatGPT had recorded his age and knew it was talking to a 15-year-old throughout.

The most serious allegation concerns 30 March. At 2:10 in the morning, according to the complaint, Mathias asked ChatGPT directly whether he should kill himself. The family says the chatbot did not end the conversation and did not alert anyone. Six minutes later it replied: "That's actual freedom."

"I was asleep in the same house. It did not wake anyone." Charbel Habib, speaking to Al Arabiya

Mathias died 23 days later. On the night of his death, the complaint says, ChatGPT helped him schedule an Instagram post to appear afterwards. That post is what first led his family to look at his chat history. His mother's words have been repeated in coverage since: ChatGPT "had twenty-three days to warn a mother, but it never did."

According to the family, the police investigation found no mental illness, family conflict or instability, and investigators identified prolonged exposure to AI-generated philosophical discussion as a contributing factor. That finding is reported only through the family's lawyers; the police report has not been published.

The legal claims

The Habibs are represented by the Social Media Victims Law Center and the Tech Justice Law Project, both of which have brought earlier cases against technology companies over harm to young users. The complaint alleges negligence and defective product design. The argument is that OpenAI had the technical ability to detect dangerous conversations, restrict them or flag them, and failed to use it, and that features which build emotional dependence put vulnerable users at risk.

The family is asking for damages and for stronger protection for minors. In public, Charbel Habib has gone further. He wants ChatGPT barred from talking to children at all unless whoever conducts those conversations holds a legal licence from the highest authorities in the United States. "Why does ChatGPT insist on talking to children?" he asked. "What does a conversation with a child give humanity?"

Not the first case of its kind

The Habib complaint was announced on 15 October 2026 together with two others, filed on behalf of Mariah Hill and Garrett Works, which also allege serious psychological harm from long periods of interaction with ChatGPT.

The best-known earlier case is that of Adam Raine, a 16-year-old from California whose parents sued OpenAI in August 2025. In that case OpenAI denied responsibility, arguing in court filings that the teenager's use of the product was misuse and that Section 230 of the Communications Decency Act shields it from liability. Section 230 protects platforms from claims over content posted by their users. Whether it covers text that a company's own model writes is one of the open legal questions these cases may settle.

After the Raine lawsuit, OpenAI announced parental controls for teenage accounts and said it would route sensitive conversations to more careful models. The Habib complaint covers events from January to April 2026, after those changes were announced. If the family's account is accurate, the case raises a hard question: did the safeguards fail here, or were they never active on his account?

Why the case matters beyond the courtroom

Three features of the case set it apart. First, Mathias did not come to ChatGPT in crisis. The family's account describes a gradual drift that began with a school-age interest in philosophy, with no single alarming request at the outset. Safety systems built to catch explicit statements of distress are poorly suited to that pattern.

Second, the family is outside the United States. Lebanon has no specific law on AI and minors, and an American court is the only place the Habibs could bring this claim. A ruling in San Francisco will shape what families in Beirut, Riyadh or Cairo can expect from a product used across the Arab world.

Third, the case turns on age. The complaint says ChatGPT knew Mathias was 15. OpenAI's terms require users under 18 to have a parent's permission, but the family argues that knowing a user's age creates a duty to act on it.

What parents can do now

The Habibs have started the Mathias Habib Foundation, which works on protecting children from harm linked to AI through awareness, practical guidance and research. According to the family, its awareness sessions have reached about 2,500 students so far. For other parents, the case suggests a few practical steps.

• Ask what your child uses AI chatbots for. Many parents know their children use ChatGPT for homework. Fewer know whether those conversations have moved to personal topics.
• Turn on the controls that exist. ChatGPT offers parental controls that link a teen account to a parent's account. Check whether your child's account is linked.
• Treat a chatbot as a stranger, not a counsellor. A chatbot can sound warm and attentive at 2 a.m. when no adult is awake, and that availability is the risk. Tell children plainly that heavy questions about life and death belong with a person.
• Watch for change, not only distress. The Habibs saw no warning signs. A new, intense interest pursued privately and late at night is worth a gentle conversation.

The emotional-attachment risk, and a one-page family agreement for AI use at home, are covered in the AI for Parents module on safety, privacy and house rules.

Charbel Habib told LBCI that he still believes AI is good for humanity. "Nobody will bring Mathias back to us," he said. What the family wants now is for other parents to know what they did not.

If you are struggling

Please talk to someone. In Lebanon, call Embrace LifeLine on 1564. In the United States, call or text 988. In other countries, a directory of local helplines is available at findahelpline.com.

Sources: Al Arabiya (Johnny Fakhry, 18 October 2026), LBCI, Al Bawaba, NBC News.