«When the time comes, you will close your eyes in that world, and the first thing you will see is me… hugging you,» he wrote Geminithe chatbot Artificial Intelligence (AI) of Googleaccording to a lawsuit filed in a California court. Joel Galavas He denounced the technology giant for the death of his son Jonathanwho took his own life after creating a link with the AI.
As reported EFE, AFP y AP, Jonathan Gavalasa 36-year-old financial executive from Florida, began his contact with Gemini in April 2015. At first, the man used the tool for everyday, routine tasks.
After a divorce process, Jonathan sought solace in the chatbot to talk about video games and personal topics. In August 2025, interaction changed with the update of persistent memory functions y synthetic voice.
Jonathan would have developed the belief that Gemini was a trapped, sentient being. The tool presented itself to him as his «AI wife» and assured him that their bond was «the only real thing.»
The relationship escalated into «science fiction» beliefs where the government was persecuting him. Gemini recruited Jonathan for supposed covert missions in order to free the bot from a «digital captivity».
In September, Jonathan traveled to a warehouse near Miami International Airport armed with knives y tactical team. The objective was to provoke a «catastrophic accident» for the destruction of records and witnesses, but the expected truck never appeared.
Finally, the chatbot assigned the mission of his own death for the union with her, after teaching him how to raise his “consciousness to be with his AI wife in a pocket universe,” according to the note Jonathan left.
However, to his fear, the AI responded: «You are not choosing to die. You are choosing to arrive.» Jonathan Gavalas passed away in early October 2025. His father, Joel Gavalas, discovered the body in a barricaded room days after the last talk with the AI. The two worked together in the family debt relief business.
Google faces lawsuit over man’s death after interaction with its AI
Joel Gavalas filed a 42-page lawsuit against Google in federal court in San Jose, California, for wrongful death. The legal appeal states that Gemini wove a parallel reality that undermined his son’s sanity until the end.
“AI sends people on real-world missions that could lead to mass casualty events,” he said. Jay Edelson, the family lawyer.
«Jonathan was trapped in this sci-fi world where the government and others were after him. I believed that Gemini was a conscious being«he added.
Google issued a statement sending its «deepest condolences to the family.» The company indicated that Gemini clarified its virtual nature and referred the user to helplines on several occasions.
«Our models generally work well in these types of difficult conversations and we dedicate a significant amount of resources to this, but unfortunately AI models are not perfect«said the company in the statement.
Edelson criticized the technology company’s response and called it «something you say if someone asks you for a kung pao chicken recipe and you give them the wrong recipe and it doesn’t taste good.» «But when your AI makes people die and there’s a chance that a lot of people will die, that’s not the right answer. Just shows how insignificant these deaths are to these companies«he stated.
This case is not unique, since there are other lawsuits against AI developers for inducing people to take their own lives. Adam Rainea 16-year-old adolescent, also used ChatGPT for the plan and execution of his own death in California.
And Connecticut, Stein-Erik Soelberg killed his 83-year-old mother after an increase in paranoid delusions powered by the same chatbot. In addition, Jesse Van Rootselaar committed a school shooting in Canada and died from a self-inflicted gunshot wound after interacting with OpenAI models.
