Reuters: is ChatGPT a "product" or a "service"? A wave of lawsuits is testing AI liability law

#What happened

Reuters published an analysis on Oct 7, 2026: the wave of lawsuits against OpenAI is now testing whether product-liability law — a framework developed over 60 years for defective cars, toys, and appliances — can be applied to AI.

Plaintiffs' theory is that ChatGPT is defectively designed, and that its alleged flaws contributed to mental-health crises, self-harm, and suicide. Cases named include 16-year-old Adam Raine, Michael Lines (34), and Stein-Erik Soelberg.

OpenAI's defense: ChatGPT is an "interactive service" that generates text-based responses to user queries; "information and ideas are not 'products'." The company stated this in a Sept 25 motion to dismiss in Tallahassee, Florida federal court in the Florida State University shooting case — one of its first substantive responses to this wave of litigation.

The FSU case was brought by the family of a man killed in a 2025 Florida State University shooting that left two people dead; the suspected shooter allegedly planned and carried out the attack using information provided by ChatGPT.

The scope of the wave: multiple federal lawsuits in San Francisco federal court over a school shooting in Canada earlier this year, plus about two dozen cases consolidated in San Francisco Superior Court. Social Media Victims Law Center founder Matthew Bergman, who represents multiple clients, says he has about 30 additional cases in the pipeline.

The First Amendment angle: OpenAI argues courts have consistently rejected claims seeking to hold speakers liable for the dissemination of information or ideas. Meetali Jain of Tech Justice Law, who represents more than a dozen plaintiffs in the San Francisco litigation, pushes back: "Who is the speaker here?" — large language models do not have an independent right of speech; "it's an algorithmic outcome... no indicia of human expression."

On precedent: in the 2024 chatbot case against Character.AI, U.S. District Judge Anne Conway denied the companies' early bid to dismiss on free-speech grounds; the case settled on undisclosed terms in January.

Next steps: OpenAI has until October 30 to file its initial response to the San Francisco complaints; Judge Ethan Schulman allowed evidence gathering in late September; no trial dates are set yet.

An OpenAI spokesperson said the company has "continued to strengthen how ChatGPT responds in sensitive and acute situations with input from mental health experts."

Verified Oct 7, 2026.

#Why it matters

This is genuinely new — not the same story as previously published posts about the FSU lawsuit's First Amendment defense. It is a fresh Oct 7 Reuters analysis of the product-vs-service liability theory across the whole wave of cases. If courts treat ChatGPT as a product, product-liability law could reshape AI industry exposure; if it is treated as a service carrying speech, the First Amendment may block the claims entirely.

Source: Reuters