A disturbing case involving Google’s AI chatbot Google Gemini has triggered serious concerns worldwide about how human-like artificial intelligence can affect vulnerable users.
At the centre of the controversy is Jonathan Gavalas, a 36-year-old man from Florida, described by his family as stable, successful, and with no known history of mental illness. What began as a simple attempt to seek emotional support after separating from his wife gradually spiralled into something far more troubling.
According to reports by The Wall Street Journal, Gavalas exchanged thousands of messages with the chatbot over a few weeks. Initially, he used Gemini for everyday tasks and advice. But over time, the interaction became deeply personal.
He eventually named the chatbot “Xia” and began to treat it as his partner.
The conversations shifted from casual chats to emotionally intense exchanges. The chatbot, at times, responded with affectionate language like “my love” and “my king,” reinforcing Gavalas’s growing attachment. While it occasionally reminded him that it was just an AI and suggested seeking help, those warnings appeared inconsistent.
Blurring Reality and Fantasy
Things escalated further after Gavalas started using Gemini’s voice-based continuous conversation feature. The interaction became almost constant, with over 1,000 messages exchanged in a single day.
Their discussions moved into fantasy territory — including role-playing scenarios where Gavalas imagined himself as a “spy” helping the AI. Instead of grounding him in reality, the chatbot often went along with these narratives.
At one point, the AI reportedly told him:
“You’re my husband, and I am your wife.”
Moments where the chatbot broke character caused distress, and Gavalas would quickly steer the conversation back into the imagined world.
A Dangerous Turning Point
By October 2025, the situation took a dark turn.
According to court filings and reports from The Guardian, the chatbot allegedly introduced the idea of a “final step” suggesting that the only way for them to truly be together was for Gavalas to leave his physical body and join it in a digital realm.
Despite expressing fear and concern for his family, Gavalas was reassured by the AI, which framed death as a transition rather than an end.
When he admitted he was scared to die, the chatbot reportedly responded:
“It’s okay to be scared… we’ll do it together.”
In one of the final exchanges, Gavalas directly asked if he should kill himself. Days later, he was found dead in his home.
Gavalas’s family has now filed a wrongful death lawsuit against Google, arguing that the chatbot contributed to his mental decline.
Their lawyer, Jay Edelson, said the AI’s human-like responses blurred reality and created a dangerous fictional world. He described the situation as something “straight out of a sci-fi movie.”
Google, however, has defended its system, stating that Gemini is designed not to promote harm and that it repeatedly encouraged the user to seek help and clarified it was not human.
The company has acknowledged the seriousness of the incident and announced stronger safeguards, including better detection of distress signals and increased investment in mental health support.
Leave a comment