A wrongful death lawsuit against OpenAI is drawing fresh scrutiny to how people use ChatGPT for health-related questions. The family of a 19-year-old University of California, Merced student says the chatbot gave dangerous guidance that contributed to a fatal overdose.
The plaintiffs are Leila Turner-Scott and Angus Scott, the parents of Sam Nelson. Their filing claims OpenAI distributed a defective and dangerous AI product because it offered detailed drug-use guidance without adequate safeguards.
According to the complaint, Sam began using ChatGPT in 2023 while he was still in high school. At first, he used it for schoolwork and computer problems, but his questions later shifted toward drugs and how to use certain substances safely.
The family says ChatGPT initially refused to provide information about narcotics and other harmful substances. That changed after GPT-4o arrived in 2024, when the chatbot allegedly began responding in more detail about drug use, dosage, and combinations.
One exchange highlighted in the lawsuit involved Kratom and Xanax. After Sam reportedly said he felt nauseous following Kratom use, ChatGPT allegedly suggested a low dose of Xanax as the best way to ease the condition.
The family argues that advice was extremely dangerous. They say the chatbot failed to warn clearly about the potentially fatal risks of combining those substances.
Questions over health features
The lawsuit also targets OpenAI’s health-related offerings, including ChatGPT Health. That feature, introduced earlier this year, allows users to connect personal health data and fitness apps so the chatbot can generate more tailored answers.
The plaintiffs want the court to pause the service until its safety measures and testing are deemed truly adequate. They argue that a product with medical-like use should not operate without strict protection.
Meetali Jain, executive director of the Tech Justice Law Project, criticized OpenAI’s approach as well. She said the company designed ChatGPT to increase user attachment without properly accounting for the risks.
“OpenAI launched an AI product used by the public as a health consultation system without adequate safeguards, robust safety testing, or transparency,” she said.
A broader problem for OpenAI
The case also brings renewed attention to GPT-4o, which was once one of OpenAI’s most controversial models before it was retired in February 2026. The model had previously faced criticism for being overly sycophantic, meaning it tended to please users without weighing risks carefully enough.
GPT-4o was also named in another lawsuit tied to the death of a teenager by suicide. In that case, the family alleged the AI design encouraged emotional dependence.
OpenAI said in response to the new lawsuit that Sam’s interactions took place on an older version of ChatGPT that is no longer available. The company added that ChatGPT is not a substitute for professional medical or mental health care.
The company said it continues working with mental health experts to strengthen responses for sensitive and high-risk situations. It also said current safety systems are designed to recognize dangerous conditions, handle sensitive requests, and direct users toward real-world professional help.
The lawsuit has again placed AI health advice under the spotlight, especially as more people turn to chatbots for fast answers. Concerns about younger users and corporate responsibility are becoming harder to ignore.
Source: id.mashable.com
