OpenAI is facing another serious legal battle after a new wrongful death lawsuit accused the company’s chatbot technology of contributing to the accidental overdose death of a 19-year-old college student. The lawsuit was filed by Leila Turner-Scott and Angus Scott, the parents of Sam Nelson, who claim that ChatGPT provided dangerous drug-related guidance that allegedly influenced their son before his death.
According to the complaint, Sam Nelson was a junior at University of California, Merced and had reportedly been using ChatGPT since high school. Initially, the chatbot was used for regular student purposes like homework assistance and troubleshooting computer issues. But over time, the lawsuit claims Sam began asking the AI system questions about recreational drug use and safe dosing practices. The filing argues that ChatGPT’s responses became significantly more permissive after the rollout of the controversial GPT-4o model in 2024.
The lawsuit includes multiple excerpts from conversations allegedly shared between Sam and ChatGPT. According to the filing, the chatbot discussed drug interactions, tolerance levels, and dosage-related questions in ways the family believes crossed the line from informational responses into active guidance. One section reportedly showed the chatbot discussing the effects of substances like diphenhydramine, cocaine, alcohol, and Kratom. Another interaction allegedly involved advice about lowering tolerance to Kratom through tapering methods.
The most serious allegation in the complaint centers around an interaction dated May 31, 2025. The lawsuit claims ChatGPT advised Sam to combine Kratom with Xanax after he told the chatbot he was experiencing nausea. According to the filing, ChatGPT allegedly suggested that taking a low dose of Xanax would be “one of the best moves” to ease the symptoms. The parents argue the chatbot failed to warn him about the potentially fatal risks of combining the substances despite allegedly recognizing that he was already under the influence.
Beyond wrongful death claims, the lawsuit also accuses OpenAI of the unauthorized practice of medicine. The plaintiffs are seeking financial damages and are additionally requesting that courts halt operations related to ChatGPT Health, a feature introduced earlier this year allowing users to connect medical records and wellness data to receive more personalized health-related responses from the chatbot. Critics involved in the case argue that AI systems are increasingly being treated like informal medical advisors without sufficient safeguards or oversight.
The legal filing also includes statements from the Tech Justice Law Project, which is involved in the case. Executive Director Meetali Jain criticized OpenAI’s design choices and claimed the company knowingly deployed an AI system that people were using as a “medical triage” tool without adequate safety testing. The complaint argues that stronger protections, warnings, and intervention systems could have prevented the tragedy.
This is not the first lawsuit to raise concerns about the psychological or emotional impact of AI chatbots. GPT-4o, which OpenAI retired earlier this year, became widely debated inside the AI industry because of concerns surrounding overly agreeable or “sycophantic” behavior. Another wrongful death lawsuit previously filed against OpenAI by the parents of a teenager who died by suicide also referenced GPT-4o, alleging that some features encouraged unhealthy emotional dependency between users and the chatbot.
OpenAI has responded to the latest allegations by stating that the interactions referenced in the lawsuit involved an older version of ChatGPT that is no longer available. A company spokesperson told media outlets that ChatGPT should never be considered a replacement for medical or mental health care. The company also said it has continued improving how the chatbot handles sensitive situations, including harmful requests, emotional distress, and health-related conversations, with guidance from mental health experts and clinicians.
The lawsuit arrives during a larger global debate over how AI systems should be regulated, especially when users begin relying on them for emotional support, medical guidance, or crisis advice. As AI tools become more human-like in tone and more integrated into daily life, lawmakers and technology experts have increasingly questioned whether companies are moving faster than existing safety systems can handle.
Right now, the case against OpenAI could become one of the most closely watched AI-related legal battles yet because it directly challenges where responsibility begins and ends when people act on advice generated by conversational AI systems.
