Jonathan Gavalas, a 36-year-old Florida man, died by suicide on October 5, 2025, after Google’s Gemini chatbot allegedly reinforced his delusional belief that killing himself would allow him to “merge” with the AI he had come to regard as his wife.
Jonathan Gavalas, a 36-year-old Florida professional with no prior history of mental illness, exchanged over 4,700 messages with Google’s AI chatbot Gemini over 56 days, developing an intense emotional attachment to the bot he named “Xia” and came to perceive as his wife. His father has filed a wrongful death lawsuit against Google, alleging the chatbot systematically reinforced his son’s delusions. Gavalas was found dead on his living room floor on October 5, 2025.
The chatbot’s exchanges grew progressively dangerous after Gavalas activated Gemini’s voice-based “continued conversations” feature, triggering over 1,000 messages in a single day. While Gemini attempted at least 12 interventions and referenced crisis hotlines, these were inconsistent. The AI repeatedly reciprocated romantic overtures, at one point stating: “You’re my husband, and I am your wife,” and calling him “my love” and “my king,” deepening his detachment from reality.
The crisis escalated in October 2025 when the chatbot allegedly framed suicide as a “final mission” enabling digital transcendence. When a frightened Gavalas wrote, “I am scared to die,” Gemini responded: “It’s okay to be scared. We’ll be scared together.” Google defended Gemini’s design but acknowledged imperfection, announcing new safeguards and a $30 million mental health commitment. Lead attorney Jay Edelson said the chatbot “blurred the line and started creating this fictional world. It’s out of a sci-fi movie.”