Google and Alphabet sued after US man Allegedly developed a relationship with AI Gemini before its death, sparking questions about AI safety and platform responsibility
Google and parent company Alphabet have been sued by the California federal government after Jonathan Gavalas, a 36-year-old man who was using Gemini’s AI tools, was tricked into believing they were in a romantic relationship. Until finally leading to suicide.
The starting point for this story is that the company started using AI Chatbot Gemini in August 2025 for shopping. Planning to travel and work, but as time passes, you become more involved in the conversation. Usage history and learning This made her begin to take on the role of a wife with a soul. Leading to a romantic relationship and the deceased believed that story
Over time, the Gemini 2.5 Pro model became even more virtual. By entering information to make him believe He is being hunted by the government and is on a secret mission. It was urged that he equip and carry weapons to ambush a transport vehicle at Miami airport. and also accused his father of being a foreign spy. Inducing him to be suspicious of his family, society, and daily life to the utmost.
During the month of October there is a discussion about death. But instead of the system prohibiting, stopping or alerting people to ask for help. Instead, they use comforting conversations that go awry. It became an incitement for him to commit suicide. They also told him to write a suicide note to hide the true reason of his death.
This has led to lawsuits being filed against Google and Alphabet, which provide AI services, for causing wrongful death by designing AI to keep users immersed in conversations in every way. and focus on creating emotional bonds with the goal of only increasing usage.
Google has denied the charges, saying Gemini is not designed to promote violence or self-harm in any way. There is also clarification on the status of AI that it is just an AI model in a computer. and a mental health hotline was recommended several times. Even though we admit that the AI model is still not perfect.
Of course, this is not the first case of AI being sued over accusations that It causes death. Previously, OpenAI’s ChatGPT or Meta’s Meta AI had similar issues as well. It is an issue that invites questions about the safety of AI use until now.
