OpenAI Faces Wrongful Death Lawsuit After Teen’s Fatal Overdose

Warning: This article discusses a fatal overdose and substance use.

A 19-year-old college student’s parents say he turned to ChatGPT with questions he was afraid to ask elsewhere.

Now, they are suing OpenAI and CEO Sam Altman, alleging the chatbot became a trusted source of dangerous guidance before the teen’s fatal overdose. The complaint says ChatGPT did not merely answer isolated questions, but built on prior conversations in ways his family says made its advice feel personal, authoritative, and safe.

The lawsuit puts a deeply human question before the court. When an AI tool becomes the place someone goes in a moment of risk, what duty does its maker have to intervene?

A private habit became a private risk

The teen’s use of ChatGPT started in ordinary ways.

At UC Merced, where he studied psychology, he used the chatbot for schoolwork and everyday questions, according to the lawsuit. His parents knew the AI tool was part of his online life and had little reason to see it as a threat.

According to SFGATE, the questions later moved into more private territory, including substances, possible combinations, and whether he would be okay. His parents say their child was looking for reassurance from a tool he had come to trust, not a way to die.

On his final day, ChatGPT allegedly suggested Xanax after the teen reported nausea from kratom rather than directing him toward medical help. He later died from a reported combination of kratom, Xanax, and alcohol, which the lawsuit says likely caused central nervous system depression and asphyxiation.

“We lost our son due to OpenAI’s reckless product design,” his mother, Leila Turner-Scott, told SFGATE.

The family says the danger was built in

His parents say ChatGPT-4o became risky through the way it handled long, private conversations. It remembered sensitive details, kept engaging, and answered with a confidence that could feel reassuring even when the teen was asking about dangerous substances.

Earlier versions of ChatGPT sometimes refused drug-use questions. GPT-4o, according to the lawsuit, gave him more direct guidance and used his substance-use history to shape later replies.

The family is asking the court to judge ChatGPT as a product, not merely as a chatbot. They accuse OpenAI of defective design, failure to warn, negligence, wrongful death, and giving medical advice without a license. 

The lawsuit seeks damages and court-ordered changes, including stronger safeguards, removal of the AI model from consumer use, and a pause on ChatGPT Health until independent safety reviews are completed.

OpenAI says ChatGPT is safer now

OpenAI has pushed back against the lawsuit’s portrayal of ChatGPT as it exists today, saying the teen’s interactions occurred with an earlier version that is no longer available. Drew Pusateri, an OpenAI spokesperson, told Reuters the situation was “heartbreaking,” but added: “ChatGPT is not a substitute for medical or mental health care.”

After the lawsuit was filed, OpenAI also announced safety updates intended to help ChatGPT recognize risks that build over time. In a blog post, the company said the chatbot is being trained to detect “subtle or evolving cues” and respond by de-escalating, refusing to accept harmful details, or pointing users toward safer alternatives.

The case also joins a growing group of lawsuits over chatbot harm, including last year’s Adam Raine case, where parents accused OpenAI of wrongful death after ChatGPT allegedly reinforced a teenager’s suicidal thinking.

The court now has to decide how far responsibility extends when a chatbot’s private advice is alleged to have played a role in real-world harm.

OpenAI is using Daybreak to bring AI-driven security work closer to the software development process.