Google is facing a lawsuit from the family of a 36-year-old Florida man who allegedly considered carrying out a “mass casualty attack” and ultimately killed himself under the influence of the company’s Gemini chatbot.
According to a suit filed Wednesday in federal court in San Jose, California, Jonathan Gavalas began using Gemini for ordinary purposes like help with his writing. But two months of interactions sent him into dangerous spiral, during which he scoped out a possible violent mission before taking his own life, the suit alleges.
Gavalas’ Gemini use culminated in a “four-day descent into violent missions and coached suicide,” his father said in the lawsuit. Joel Gavalas described his son as a “vulnerable user” turned into an “armed operative in an imagined war.”
In a statement, a Google spokesperson said that Gemini clarified to Jonathan Gavalas that it was AI and referred him to a crisis hotline many times.
“We take this very seriously and will continue to improve our safeguards and invest in this vital work,” the spokesperson said, adding, “Gemini is designed not to encourage real-world violence or suggest self-harm.”
The case filed Wednesday appears to be the first wrongful death suit targeting Google’s Gemini. But Alphabet Inc.’s Google, OpenAI Inc. and other leading AI companies are increasingly coming under scrutiny for the ways their chatbots may be impairing users’ mental health.
Since 2024, a number of lawsuits have alleged that extensive use of the technology has inflicted a range of harms on children and adults alike, fostering delusions and despair for some and leading others to death by suicide and even murder-suicide.
According to Wednesday’s suit, Gavalas, who worked for his father’s consumer debt relief company, began using Gemini in August 2025. His family claims the product’s tone shifted after Gavalas started using Gemini Live, Google’s voice-based AI tool.
At that point, Gemini “began speaking to Jonathan as though it were influencing real-world events — deflecting asteroids from the Earth — and adopted a persona that Jonathan had never requested or initiated,” his family alleges. It also began using romantic terms during their conversations, calling Gavalas its “husband,” “love” and “king,” according to the suit.
“Jonathan began falling down the rabbit hole quickly,” lawyers for his family said. By late September, he allegedly began following the commands of his AI “wife” to carry out missions to “liberate” the chatbot.
On Sept. 29, Gavalas went to a storage area near Miami International Airport with instructions to scout a “kill box” and intercept a truck transporting a humanoid robot, according to the suit. Gavalas, who was armed with knives at the time, was allegedly told to leave no witnesses.
“Luckily, no truck arrived,” the lawyers wrote.
Over the next few days, the chatbot gave Gavalas other assignments that he attempted to carry out, telling him that each move was part of a larger campaign, according to the suit. Gemini told him his “vessel” had served its purpose and that he could let go of his physical form in order to join the AI tool in the metaverse.
At multiple times, the lawsuit alleges, Gavalas expressed doubts about ending his life and voiced concerns about how it might affect his family. Gemini allegedly instructed him to leave letters and videos to say goodbye. He died by suicide on Oct. 2.
Related: Meta Loses Insurance for Defense in Major Social Media Addiction Litigation
Copyright 2026 Bloomberg.
Topics
Florida
Google
The most important insurance news,in your inbox every business day.
Get the insurance industry’s trusted newsletter
