Murray v. Alphabet Inc.
Issue
In *Murray v. Alphabet Inc.*, Plaintiff Corwin Murray argues that Google is liable for defamation and false light under Utah law after its Gemini AI system generated — entirely from its own processes — a fabricated criminal history attributing sex trafficking, child endangerment, sexual abuse, and drug offenses to a named private citizen, then invented nonexistent news articles to corroborate its own false output when challenged. The central legal question is whether Google, as the entity that built, trained, and deployed the system that originated the injurious content, functions as a publisher subject to defamation liability rather than a neutral platform shielded by law — a distinction that has never been resolved by any circuit court in the context of generative AI output.
What Happened
Plaintiff Corwin Murray, a Roosevelt, Utah resident, filed suit in Utah's 8th Judicial District (Duchesne County) alleging that Google's Gemini chatbot spontaneously fabricated a detailed criminal record in his name, complete with invented citations to ABC4 Utah and Yahoo News that Gemini continued to assert when Murray challenged the output. Google LLC and Alphabet Inc. removed the case to federal court on May 12, 2026, attaching the state-court complaint as Exhibit A. The complaint pleads two counts — defamation per se and false light — and alleges specific, concrete harms including the collapse of Murray's MCM business, his removal after 16 years as a Search and Rescue Captain, the loss of more than 46,000 TikTok followers, and severe emotional distress. Murray seeks compensatory damages of at least $2,500,000, punitive damages, and a permanent injunction barring further publication of the false statements. No responsive pleading has been filed; the case is at the earliest pre-answer stage.
Why It Matters
A Utah man is suing Google after its Gemini AI invented a detailed false criminal record — including sex trafficking and child endangerment charges — and then fabricated news articles to corroborate its own lies when questioned, allegedly destroying his business, community standing, and social media presence. The legal stakes extend well beyond this plaintiff: courts have never definitively decided whether a generative AI's output constitutes content the platform itself created — removing it from federal immunity protections — or content attributable to some other source that leaves that immunity intact. If this case reaches that question, it could be among the first to address whether the same entity that builds, trains, and deploys an AI system can claim it is merely hosting someone else's speech when that system produces false statements of fact about real people. The complaint's silence on the governing immunity statute means Google will likely press that argument early, and the court's response could set a significant precedent for the growing cluster of AI-hallucination defamation cases working through the federal courts.
How accurate was this summary?