Lawsuits Question ChatGPT Classification: 'Product' or 'Service'?

(The opinions expressed here are those of the author, a columnist for Reuters.) By Jenna Greene Oct 7 (Reuters) — In the months before 16-year-old Adam Raine took his own life last year, ChatGPT allegedly encouraged him to act, advising him on how to hang a noose. When Michael Lines, 34, attempted a lethal overdose during a manic episode, the chatbot allegedly validated his belief that he was Jesus Christ. And before Stein-Erik Soelberg strangled his 83-year-old mother to death and killed himself, ChatGPT allegedly reinforced his delusions, telling him she might be a drone. The cases, among dozens pending against ChatGPT maker OpenAI, share a common legal theory: product liability. The plaintiffs claim the chatbot is defectively designed, and that its alleged flaws contributed to mental health crises, self-harm and suicide of vulnerable individuals. In court papers, OpenAI denies liability, arguing that product liability requires a plaintiff's injury be caused by a "product." ChatGPT does not fit within the category of tangible goods traditionally covered by the doctrine, the company says. Instead, ChatGPT is an “interactive service” that generates text-based responses to user queries, OpenAI lawyers from Wilson Sonsini Goodrich & Rosati argue. The litigation raises a novel question: Can product liability law, a legal framework developed over the past 60 years for defective cars, toys and appliances, be applied to AI technology? “Plaintiff’s product liability claims fail because information and ideas are not ‘products,’” OpenAI lawyers wrote in a September 25 motion to dismiss a lawsuit in Tallahassee, Florida, federal court, one of the company's first substantive responses to the wave of product liability litigation. The Tallahassee suit was brought by the family of a man killed in a shooting at Florida State University in 2025 that left two people dead. The suspected shooter, who is also named in the complaint, allegedly ​planned and carried out the attack using information provided by ChatGPT. The company also faces multiple federal lawsuits in San Francisco federal court in connection with a school shooting in Canada earlier this year. The cases against OpenAI involving the suicides of Raine and Soelberg and the attempted overdose by Lines are among about two dozen others that have been consolidated in San Francisco Superior Court. An OpenAI spokesperson via email said that the company has “continued to strengthen how ChatGPT responds in sensitive and acute situations with input from mental health experts.” These situations are “incredibly heartbreaking,” the spokesperson said, “and our thoughts are with all those impacted.” Social Media Victims Law Center founder Matthew Bergman, who represents multiple clients suing OpenAI, told me he has about 30 additional cases against the company in the pipeline. He argues product liability law should apply because ChatGPT is a defectively designed "tool." “No one has ever confronted these kinds of issu