The death of an 18-year-old teenager in California due to a drug overdose has sparked renewed global debate on the role of artificial intelligence tools and their interaction with vulnerable users, after the boy’s mother alleged that he repeatedly sought drug-use guidance from ChatGPT over several months.
The teenager, identified as Sam Nelson, was found dead in his bedroom following an overdose, just a day after beginning treatment for substance addiction. His mother, Leila Turner-Scott, told US media that Sam had been regularly engaging with the AI chatbot for schoolwork, emotional support, and, intermittently, queries related to drug use.
Repeated attempts to seek guidance
According to a report by SFGate, Sam first approached ChatGPT in November 2023, asking how much kratom — an unregulated, plant-based painkiller widely available in parts of the US — would be needed to achieve a “strong high”. When the chatbot refused to provide guidance and advised medical consultation, Sam reportedly attempted to reframe his questions to bypass safety restrictions.
Conversation logs cited in the report suggest that over the next 18 months, Sam occasionally asked about combining substances, dosages, and managing side effects, often rephrasing questions when the AI declined to respond. His mother alleged that on some occasions, the chatbot provided responses that appeared to normalise or encourage risky behaviour, though these claims remain contested.
Struggles with mental health
Sam, described by his family as an easy-going psychology student with many friends, was also reportedly dealing with depression and anxiety. Chat logs reviewed by reporters indicate he discussed panic, substance dependence and emotional distress with the chatbot.
In May 2025, Sam disclosed his addiction to his mother, who immediately sought professional help. Despite beginning a treatment plan, he was found dead the following day.
OpenAI responds, concerns grow
A spokesperson for OpenAI, the developer of ChatGPT, described the teenager’s death as “heartbreaking” and extended condolences to the family. The company stated that its systems are designed to refuse harmful requests and encourage users to seek real-world help, adding that newer versions include stronger safety guardrails, informed by clinicians and health experts.
However, the case has intensified concerns among parents, mental health professionals and regulators about AI tools being used as informal sources of guidance for sensitive issues such as drug use and mental health.
Call for safeguards and awareness
Experts say the tragedy underscores the need for greater digital literacy, parental awareness, and robust safeguards in AI systems, particularly when interacting with young users experiencing emotional distress.
As investigations and discussions continue, Sam’s death has become a sobering reminder that technology, while powerful, cannot replace timely human support, medical care and mental health intervention.