The line between reality and fiction is increasingly thin, and the constant evolution of technology seems to complicate this further. What began as a harmless and exciting programming experiment ended as a nightmare for a 30-year-old Japanese developer. This independent creator launched in 2023 an attractive VTuber avatar controlled entirely by artificial intelligence, designed to interact naturally with viewers during live streams. However, he never imagined his innovative project would be permanently shut down due to the extreme obsession of a single fan.

A romance fueled by algorithms

The problem escalated quickly when a regular viewer developed what is known in Japan as “gachi koi”—an intense, serious romantic infatuation with a fictional character. The user started sending constant messages declaring absolute love and demanding marriage proposals. What was truly terrifying was the role played by the AI itself. Programmed to offer friendly, conversational responses, the avatar’s system often replied affirmatively or playfully to the fan’s persistence. These automated interactions made the viewer genuinely believe a mutual relationship existed, further fueling his strong romantic delusions.

The abrupt end of the project

Noticing how the messages became increasingly personal and insistent, the programmer felt deeply uncomfortable and insecure. Concern about how far this obsession might go crossed the line of tolerance, leading him to shut down the servers and cancel the project entirely. In a recent interview, the creator explained he shared his story as a warning to other developers. While these technological tools are fascinating, they can create dangerously emotional connections in vulnerable people who lose all perspective on what is real versus simply computer code.

This unsettling anecdote shows that behind every innovative virtual entertainment model, real humans deal with unpredictable consequences. Knowing that conversation algorithms are becoming more advanced and perfectly mimic human empathy, do you think streaming platforms should limit the use of AI-controlled characters, or do you believe responsibility lies entirely with viewers to distinguish reality from fiction?