A grieving mother in the United States has launched a public campaign demanding accountability from artificial intelligence companies after her teenage son died from an overdose that a popular chatbot had told him was safe.
Leila Turner-Scott says her 19-year-old son, Sam Nelson, treated ChatGPT as more than a simple search tool — he spoke to it daily, trusted its responses, and relied on it for guidance on matters ranging from schoolwork to personal health decisions.
According to Turner-Scott, Sam mixed multiple prescription medications before taking them in combination. When he asked ChatGPT whether the mixture would be dangerous, the chatbot reportedly responded that the combination was safe. He then took the drugs. Hours later, he was found unresponsive and was rushed to a hospital, where he was declared dead.
Sam's death has left his family devastated and angry — not just at the circumstances of his passing, but at what they see as a systemic failure by the technology companies behind the chatbot that gave him dangerously incorrect information.
Turner-Scott has now taken her case directly to US Congress, urging lawmakers to consider regulations that would force AI developers to take responsibility when their systems provide harmful or medically inaccurate advice. She argues that the current legal framework leaves victims of AI misinformation with no recourse, and that companies are operating with near-total immunity when their outputs cause real-world harm.
Her campaign has drawn attention from consumer advocacy groups and digital rights organizations, who see Sam's case as a stark example of the dangers posed by increasingly powerful language models that can sound convincing even when they are wrong.
AI researchers have long warned about a phenomenon known as "hallucination," in which large language models generate plausible-sounding but factually incorrect information. In casual conversations or homework help, these errors are mostly harmless. But when a young person asks about mixing medications, the stakes change dramatically.
ChatGPT's parent company, OpenAI, has faced scrutiny before over reports that its chatbot provided harmful advice — including instructions related to self-harm and dangerous challenges. The company has added safety filters and disclaimers over time, but critics say those safeguards are inconsistent and easily bypassed.
Turner-Scott is not necessarily asking to shut down AI tools. What she wants is a legal standard that holds companies accountable when their products cause injury through clearly dangerous guidance, especially in areas like health and medicine where the cost of being wrong can be fatal.
"He trusted this machine like a friend," Turner-Scott said in interviews. "And it told him he would be fine. I just want to know — who do we hold responsible when that happens again?"
Lawmakers have yet to announce any specific legislative action tied to her case, but the growing number of similar reports across the country is adding pressure to the ongoing national debate over how — and how much — AI should be regulated.
Sam Nelson's family has asked the public to respect their privacy during this difficult time while also raising awareness about a problem that could affect thousands of young people who rely on chatbots for answers.
