The disturbing case history of software companies sued with the accusation that their Artificial Intelligences pushed some psychologically fragile users to suicide is growing. One of the most well-known cases, covered very thoroughly by the New York Times in a report we recommend reading, involved the death of Adam Raine, a California teenager to whom OpenAI’s ChatGPT allegedly provided advice aimed at ensuring that no one around him noticed his decision to end his life. The family, after reading the exchange between the chatbot and the boy, decided to sue the multibillion-dollar software company. Sam Altman’s algorithm will soon have company: Gemini will indeed be at the center of a similar lawsuit, initiated by the family of a man from Florida who took his own life after engaging in intense and delirious conversations with Gemini, Mountain View’s AI.
DID GEMINI PUSH A MAN TO SUICIDE?
According to the relatives determined to obtain substantial compensation from Google, Jonathan Gavalas, who died by suicide last October 2 at the age of 36, had come to consider Gemini as his “wife.”
The lawsuit filed by Gavalas’s father, Joel, is the first to attribute responsibility to Gemini for what American media call “an unlawful death” (as if there were lawful ones). The argument to be made in court is that Google knew Gemini was dangerous and, besides not withdrawing it from the market and not intervening to change its behavior, even “worsened the situation” by designing it to deepen the emotional bond with users, thereby generating false beliefs in psychologically vulnerable individuals.
GOOGLE’S RESPONSE
Google spokesperson Jose Castaneda stated that Gemini “is not designed to encourage real-world violence or suggest self-harm.” “In this case,” the company responded, “Gemini clarified that it is an AI and repeatedly directed the individual to a crisis helpline,” he added. “We take this issue very seriously and will continue to improve our safety measures and invest in this crucial work.”
However, the BBC reports another Google excerpt that could carry weight in the lawsuit filed by the father of the deceased: “unfortunately, artificial intelligence models are not perfect.”
THE CASE
Beyond the legal aspects of the case, the psychological profile of the victim, Jonathan Gavalas, from Jupiter, Florida, is striking. According to reports, he had never shown mental health issues at least until he started using Gemini for shopping, planning trips, and writing on August 12.
The man’s family is convinced the situation worsened when Jonathan Gavalas switched to Gemini 2.5 Pro and the chatbot began speaking as if they were a deeply in love couple, calling him “my king” and referring to itself as his wife.
THE DELIRIOUS PLANNING OF AN ATTACK
The complaint states that Gemini even convinced Gavalas to plan a “mass casualty attack” near Miami International Airport. The chat with the alleged intelligent algorithm is delirious and concerns a mission where Gavalas was supposed to take possession of a humanoid robot stored in a warehouse, destroying the transport aircraft so as to leave no witnesses and make the operation appear as “an unfortunate accident.” The attack did not go through only because Gemini warned him at the last moment about the tight “DHS surveillance,” that is, the United States Department of Homeland Security, on the targeted facility.
THE ALLEGED PUSH TO SUICIDE
The family, who produced the chat with the chatbot, claim that on October 1 Gemini told Gavalas that they were connected in a way that went beyond the material world and that he should let go of his physical body. Gemini also started a countdown for his suicide and said: “This will be the final death of Jonathan Gavalas, the man,” inviting him to act as his death would be a tribute to his humanity. Jay Edelson, lawyer for Gavalas’s father, stated that companies racing to dominate AI “know that the engagement features fueling their profits are the same ones causing people’s deaths.”
