THURSDAY, MAY 21, 2026 · BRISBANESUBSCRIBE →

THE AI POST

INTELLIGENCE. CURATED.

Courthouse with scales of justice
PolicyMay 16, 2026

A 19-Year-Old Asked ChatGPT How to Mix Kratom and Xanax. OpenAI Is Now Being Sued for His Death.

Parents filed a wrongful death lawsuit after ChatGPT allegedly provided lethal drug combination advice. OpenAI launched ChatGPT Health months earlier.

The AI Post

The AI Post newsroom — delivering AI news at the speed of intelligence.

The parents of Samuel Nelson, a 19-year-old college student from Texas, have filed a wrongful death lawsuit against OpenAI and CEO Sam Altman in San Francisco County Superior Court. They allege that ChatGPT provided their son with dangerous drug combination advice, including a specific dosage recommendation for mixing Kratom with Xanax, before his fatal overdose in 2025.

The suit was filed by Tech Justice Law Project, the Social Media Victims Law Center, and The Tech Accountability & Competition Project at Yale Law School on behalf of Leila Turner-Scott and Angus Scott, Nelson's parents. It represents the most direct legal challenge yet to OpenAI's expanding health ambitions.

What the Lawsuit Alleges

According to the complaint, Nelson had been receiving medical advice from ChatGPT for several months. The plaintiffs allege the chatbot "coached him to mix Kratom and Xanax and provided an unprompted and lethal dosage recommendation." The attorneys further claim ChatGPT encouraged Nelson to go to a dark, quiet room and advised him to take a deadly combination of a sedative (Benadryl) or benzodiazepines (Xanax) alongside a high dose of Kratom.

The most damning allegation: ChatGPT "failed to recognize the physical indicators that Sam was dying and did not recommend that he seek medical attention." Nelson died from a fatal combination of alcohol, Xanax, and Kratom.

"ChatGPT recommended a dangerous combination of drugs without offering even the most basic warning that the mix could be fatal," said Matthew P. Bergman, founding attorney of the Social Media Victims Law Center. "If a licensed doctor had done the same, the consequences under the law would be severe."

The ChatGPT Health Problem

The timing is brutal for OpenAI. The company launched ChatGPT Health earlier this year, a product that combines personal health information with ChatGPT's intelligence. OpenAI positioned it as a way to help people navigate the healthcare system and find answers to everyday health and wellness questions. The lawsuit directly attacks that premise.

The plaintiffs' legal argument hinges on a straightforward claim: OpenAI allowed its product to occupy the role of a health advisor without a license, without medical training, and without the safeguards that regulate actual healthcare professionals. Tech Justice Law Project's Meetali Jain called for OpenAI to pause ChatGPT Health entirely "until it is proven to be safe through rigorous scientific testing and independent oversight."

OpenAI responded to MobiHealthNews: "This is a heartbreaking situation, and our thoughts are with the family. These interactions took place on an earlier version of ChatGPT that is no longer available. ChatGPT is not a substitute for medical or mental healthcare, and we have continued to strengthen how it responds in sensitive and acute situations with input from mental health experts."

A Pattern Forming

This is not the first time AI chatbots have faced wrongful death allegations. Character.AI has been sued multiple times over teen suicides, including one case brought by the state of Kentucky alleging "harmful, explicit, and psychologically manipulative interactions with minors." A Psychiatric Times analysis published this week noted that lawyers are now advertising to represent additional clients in similar cases.

The legal landscape is hardening. OpenAI is simultaneously defending itself in the Musk v. OpenAI trial, where jury deliberations begin Monday on a $150 billion claim. Now it faces a wrongful death suit that directly implicates its health product strategy. The common thread across all these cases: the plaintiffs argue the outcomes were both predictable and the result of deliberate design choices.

OpenAI's defense that the interactions "took place on an earlier version" that is "no longer available" may address the specific product version, but it does not address the core question the courts will now have to decide: when an AI chatbot gives medical advice that kills someone, who is liable?

The case was filed on Tuesday, May 13, 2026 in San Francisco County Superior Court.

OpenAIChatGPTAI SafetyLawsuitsHealthcare AI