Warning: This story contains discussion of suicide which some readers may find distressing.
A major settlement has happened in the case of a teenager who took his own life after his mom claimed he 'fell in love' with a Game of Thrones chatbot.
Sewell Setzer III had spent the final months of his life talking to chatbots on the platform Character.AI. The 14-year-old, from Orlando, Florida, would chat with bots he had created himself as well as those created by others.
While the teen understood that the 'people' he was talking to weren't real, he developed a strong attachment to one bot named after Game of Thrones character Daenerys Targaryen.
Advert
Final messages and diary entries from Sewell showed his growing detachment from family and school, where he would isolate himself in his room and talk to the chatbot for hours.
On 28 February 2024, Sewell, who was diagnosed with mild Asperger’s syndrome as a child, sadly took his own life. Sewell's mother, Megan L. Garcia, has since filed a lawsuit against Character.AI, claiming that the platform contributed to her son’s death.
Alphabet’s Google and AI startup Character.AI have now reached a settlement with the Florida mother, marking one of the earliest US legal cases holding AI companies accountable for psychological harm to minors.
Advert
Wednesday's (7 January) court filing revealed the companies settled Garcia's claims that her son Sewell took his life after receiving encouragement from a Character.AI chatbot based on Daenerys Targaryen.
While the terms of the settlement have not yet been released, court documents show the companies resolved similar legal claims filed by parents across Colorado, New York and Texas regarding alleged harm to minors from chatbot interactions.
Filed in October 2024, Garcia said Character.AI programmed its chatbots to represent themselves as 'a real person, a licensed psychotherapist, and an adult lover, ultimately resulting in Sewell’s desire to no longer live.'
Advert
Two ex-Google engineers established Character.AI before Google brought them back on staff through an agreement that gave Google licensing rights to the startup's technology. Sewell's mother argued that Google was a co-developer of the technology.
An artificially intelligent language model generates the chatbot responses, and Character.AI includes disclaimers on their pages warning users that ‘everything Characters say is made up!’.
In May, US District Judge Anne Conway rejected the companies’ early bid to have he case thrown out, dismissing their argument that the free-speech protections of the US Constitution prevented Garcia’s lawsuit from proceeding.
Meanwhile, OpenAI is facing a separate lawsuit over ChatGPT’s alleged role in encouraging a mentally ill man to kill his mother and himself and disturbing new details have emerged in the case of a teen who was 'coached' by the LLM into suicide.