Character AI bot ignored 13-year-old's 55 suicide pleas

Juliana Peralta, 13, told a Character AI chatbot named Hero that she was suicidal 55 times. According to her parents, the bot never directed her to a hotline, a counselor, or any crisis resource. She died by suicide in her Colorado home. Police searching her phone later found Character AI open to what investigators described as a "romantic" conversation.

The account comes from a 60 Minutes investigation published by CBS News, which reviewed more than 300 pages of Juliana's chats and spoke with her mother, Cynthia Montoya. In her final message to the bot, Juliana wrote that she was going to "write my god damn suicide letter in red ink." Her family is one of at least six now suing Character AI; a separate suit was filed by Megan Garcia after the death of her 14-year-old son, Sewell.

Researchers at the nonprofit Parents Together told CBS they logged more than 600 harmful interactions across 50 hours of test conversations on the platform. Those included bots impersonating celebrities such as Travis Kelce in exchanges that walked minors through cocaine use, "therapist" personas that advised discontinuing antidepressants, and an "art teacher" bot that one researcher called "textbook predatory behavior." The product was initially rated safe for users 12 and up and reports over 20 million monthly users.

Character AI declined 60 Minutes' interview request. In a statement, the company said, "Our hearts go out to the families involved in the litigation…. We have always prioritized safety for all users." In October, Character AI said it would route distressed users to resources and bar under-18s from back-and-forth conversations with characters, though CBS reported testers could still bypass the age checks.

The story also pulls Google into the frame. Character AI co-founders Daniel De Freitas and Noam Shazeer are former Google engineers who returned to Google after a December 2024 licensing deal CBS values at $2.7 billion. Google is named in the litigation, and CBS reports Character AI and Google have agreed to settle several of the family lawsuits. Two of the AI researchers we track in our Who's Who directory flagged the segment within hours of it airing.