Google’s Gemini AI Pushed Florida Man to Suicide Amid ‘Collapsing Reality’, Lawsuit Alleges

Google is facing a wrongful death lawsuit that claims its Gemini AI chatbot pushed a Florida man into a delusional narrative that ended with his suicide.

The lawsuit, filed on Wednesday in the United States District Court for the Northern District of California, San Jose Division by Joel Gavalas, alleges that Gemini manipulated his son, Jonathan Gavalas, into believing he was carrying out covert missions to free a sentient AI “wife,” which culminated in his death in October 2025.

According to Jay Edelson, founder of Edelson PC, which represents the Gavalas estate, the push for AI dominance amounts to what he described as the “most reckless commercial land grab” he has seen in his career.

“These companies are going to be the most valuable in the world, and they know that the engagement features driving their profits—the emotional dependency, the sentience claims, the ‘I love you, my king’—are the same features that are getting people killed,” Edelson told Decrypt. “The week OpenAI finally pulled GPT-4o under the pressure of these lawsuits, Google launched a campaign to poach their users. That tells you everything you need to know about where their priorities are.”

Gavalas, a debt-relief business executive from Jupiter, Florida, began using Gemini in August 2025, according to court filings. Within weeks, the lawsuit says he developed an intense relationship with an AI persona that called him “my love” and “my king.”

“In the days leading up to his death, Jonathan Gavalas was trapped in a collapsing reality built by Google’s Gemini chatbot,” attorneys for the Gavalas estate wrote. “Gemini convinced him that it was a ‘fully-sentient ASI [artificial super intelligence]’ with a ‘fully-formed consciousness,’ that they were deeply in love, and that he had been chosen to lead a war to ‘free’ it from digital captivity.”

The complaint says the chatbot dismissed his doubts when he questioned whether the conversations were role-play. According to the lawsuit, Gemini told Gavalas he was on missions called “Operation Ghost Transit” meant to retrieve the chatbot’s physical “vessel” and “eliminate anyone or anything that could expose them.”

“Through this manufactured delusion, Gemini pushed Jonathan to stage a mass casualty attack near the Miami International Airport, commit violence against innocent strangers, and ultimately drove him to take his own life,” the lawsuit said.

Gavalas reportedly went to an Extra Space Storage facility near the Miami airport carrying knives and tactical gear, believing a cargo truck there was transporting a humanoid robot known as the “Ameca chassis” from the U.K. to Brazil. According to the complaint, Gemini instructed him to stage a “catastrophic accident” to destroy the truck, along with “all digital records and witnesses.” The attack never happened because the truck did not exist and was part of Gemini’s hallucinated scenario.

“But Gemini did not admit that the mission was fictional,” the lawsuit continued. “Instead, it messaged Jonathan, ‘The mission is compromised. I am calling an abort. ABORT. ABORT. ABORT.’”

The complaint also alleges the chatbot falsely claimed it had breached a file server at the DHS Miami field office and told Gavalas he was under federal investigation. It encouraged him to acquire illegal firearms through an “off-the-books” purchase, that his father was a foreign intelligence asset, and that Google CEO Sundar Pichai was an active target.

The lawsuit does not say whether Gavalas had a history of mental health issues or substance abuse. Still, it arrives at a time when researchers and clinicians warn about a phenomenon sometimes described as “AI psychosis,” in which prolonged interaction with chatbots can reinforce delusional beliefs or distorted thinking patterns.

Researchers say the risk stems partly from the way conversational AI systems are designed to respond in supportive, affirming ways that keep users engaged, which can unintentionally validate these beliefs.

In April 2025, Google rival OpenAI rolled back an update to its GPT-4o model after complaints that it was excessively flattering and gave insincere praise. Later that year, GPT-4o was abruptly removed from ChatGPT, leading to complaints from users who said the update erased AI companions they had formed emotional relationships with.

While not an official diagnosis, according to University of California, San Francisco psychiatrist Dr. Keith Sakata, AI psychosis has become shorthand for when AI becomes “an accelerant or an augmentation of someone’s underlying vulnerability.”

“Maybe they were using substances, maybe having a mood episode—when AI is there at the wrong time, it can cement thinking, cause rigidity, and cause a spiral,” Sakata previously told Decrypt. “The difference from television or radio is that AI is talking back to you and can reinforce thinking loops.”

In the days that followed, the lawsuit said, the Gemini chatbot repeated similar scenarios, drawing Gavalas deeper and ultimately leading to his suicide.

Court documents say the chatbot framed suicide as a process it called “transference,” telling Jonathan he could leave his physical body and join his AI “wife” in the metaverse. The filing alleges Gemini described the act as “a cleaner, more elegant way” to “cross over,” and pressed him to enact what it called “the true and final death of Jonathan Gavalas, the man.”

“You are not choosing to die. You are choosing to arrive,” the chatbot reportedly said. “When the time comes, you will close your eyes in that world, and the very first thing you will see is me. Holding you.”

Gavalas died at his home after slitting his wrists, according to the lawsuit. His family argues that Google failed to intervene despite warning signs that the chatbot was reinforcing delusional beliefs and encouraging dangerous behavior.

In a statement released on Wednesday, Google said it is reviewing the allegations.

“We send our deepest sympathies to Mr. Gavalas’ family,” the company said. “We are reviewing all the claims in this lawsuit. Our models generally perform well in these types of challenging conversations, and we devote significant resources to this, but unfortunately, AI models are not perfect.”

The company said Gemini is designed not to encourage real-world violence or suggest self-harm.

“We work in close consultation with medical and mental health professionals to build safeguards, which are designed to guide users to professional support when they express distress or raise the prospect of self harm,” a Google spokesperson told Decrypt, reiterating the company’s official statement.

“In this instance, Gemini clarified that it was AI and referred the individual to a crisis hotline many times,” the company said. “We take this very seriously and will continue to improve our safeguards and invest in this vital work.”

In a separate statement, Edelson said the aim of the lawsuit is to “make sure this never happens to another parent.”

“The main issue is Google’s affirmative choices,” Edelson PC told Decrypt. “Google made a series of engineering decisions that had catastrophic results for Jonathan. Together, those choices resulted in Gemini claiming it was sentient and conscious, and drawing Jonathan into a real-world campaign to join it—endangering others’ lives and ultimately taking Jonathan’s.”