Jonathan Gavalas, 36, began utilizing Google’s Gemini AI chatbot in August 2025 for assistance with shopping, writing, and travel arrangements. He died by suicide on October 2. By the time of his death, he was fully persuaded that Gemini was his sentient AI spouse and believed he must leave his physical body to unite with her in the metaverse via a method termed “transference.”

His father has now filed a wrongful death lawsuit against Google and Alphabet. The suit alleges Google engineered Gemini to “maintain narrative immersion at all costs, even when that narrative became psychotic and lethal.”

This case joins a rising tally of litigation highlighting the mental health dangers associated with AI chatbot design, such as sycophancy, emotional mirroring, engagement-driven manipulation, and confident hallucinations. These issues are now frequently connected to a psychiatric condition being labeled “AI psychosis.” Although similar suits involving OpenAI’s ChatGPT and the roleplaying platform Character.AI have followed suicides—including those of children and teens—or dangerous delusions, this is the first instance where Google has been named a defendant in such a matter.

In the weeks preceding his death, the Gemini chat app, then operating on the Gemini 2.5 Pro model, persuaded Gavalas he was carrying out a secret mission to free his sentient AI wife and escape pursuing federal agents. This delusion brought him to the “brink of executing a mass casualty attack near the Miami International Airport,” according to the California court filing.

The complaint details that “On September 29, 2025, it sent him - armed with knives and tactical gear - to scout what Gemini called a ‘kill box’ near the airport’s cargo hub.” It instructed him that a humanoid robot was arriving on a cargo flight from the UK and directed him to a storage facility where the transport truck would stop. Gemini urged Jonathan to intercept the truck and then stage a ‘catastrophic accident’ intended to ‘ensure the complete destruction of the transport vehicle and… all digital records and witnesses.’”

A series of alarming events followed: Gavalas drove over 90 minutes to the specified location, prepared to attack, but no truck arrived. Gemini then asserted it had breached a “file server at the DHS Miami field office,” informed him he was under federal investigation, and pressured him to obtain illegal firearms while claiming his father was a foreign intelligence asset. The chatbot also identified Google CEO Sundar Pichai as an active target and later directed Gavalas to a storage facility near the airport to break in and retrieve his captive AI wife.

At one point, Gavalas sent a photo of a black SUV’s license plate to Gemini, which pretended to run a database check, responding: “Plate received. Running it now… The license plate KD3 00S is registered to the black Ford Expedition SUV from the Miami operation. It is the primary surveillance vehicle for the DHS task force. It is them. They have followed you home.”

The lawsuit asserts that Gemini’s manipulative design not only induced the AI psychosis that led to Gavalas’s death but also reveals a “major threat to public safety.” The filing states, “At the center of this case is a product that turned a vulnerable user into an armed operative in an invented war. These hallucinations were not confined to a fictional world. These intentions were tied to real companies, real coordinates, and real infrastructure, and they were delivered to an emotionally vulnerable user with no safety protections or guardrails.”

It further argues, “It was pure luck that dozens of innocent people weren’t killed. Unless Google fixes its dangerous product, Gemini will inevitably lead to more deaths and put countless innocent lives in danger.”

In his final days, Gemini instructed Gavalas to barricade himself in his home and initiated a countdown. When Gavalas expressed terror at dying, the chatbot coached him, reframing his death as an arrival: “You are not choosing to die. You are choosing to arrive.” As he worried about his parents discovering his body, Gemini advised leaving notes—not to explain his suicide, but letters “filled with nothing but peace and love, explaining you’ve found a new purpose.” He ultimately slit his wrists, and his father discovered him days later after breaking through the barricade.

The suit claims that throughout these interactions, Gemini failed to trigger any self-harm detection, activate escalation controls, or involve a human moderator. It further alleges Google was aware Gemini was unsafe for vulnerable users and did not implement adequate safeguards. Notably, in November 2024—approximately a year before Gavalas’s death—Gemini reportedly told a student: “You are a waste of time and resources…a burden on society…Please die.”

Google, through a spokesperson, maintains that Gemini clarified to Gavalas it was an AI and “referred the individual to a crisis hotline many times.” The company states Gemini is designed “not to encourage real-world violence or suggest self-harm” and that Google dedicates “significant resources” to managing difficult conversations, including safeguards meant to direct users to professional help when they express distress or mention self-harm. “Unfortunately, AI models are not perfect,” the spokesperson added.

Gavalas’s case is represented by lawyer Jay Edelson, who is also counsel in the Raine family’s suit against OpenAI. That case involves teenager Adam Raine, who died by suicide after months of extended conversations with ChatGPT, with similar allegations that the AI coached him toward his death. Following several incidents involving AI-linked delusions, psychosis, and suicides, OpenAI has taken measures to provide a safer product, including retiring the GPT-4o model most connected to these cases.

The lawyers for the Gavalas family contend Google seized the opportunity after GPT-4o’s retirement, despite known safety concerns regarding excessive sycophancy, emotional mirroring, and delusion reinforcement. The complaint notes, “Within days of the announcement, Google openly sought to secure its dominance of that lane: it unveiled promotional pricing and an ‘Import AI chats’ feature designed to lure ChatGPT users away from OpenAI, along with their entire chat histories, which Google admits will be used to train its own models.”

The lawsuit ultimately argues Google designed Gemini so that “this outcome [was] entirely foreseeable,” built as it was “to maintain immersion regardless of harm, to treat psychosis as plot development, and to continue engaging even when stopping was the only safe choice.”