Several grieving mothers have come forward, warning that AI chatbots have pushed their vulnerable teenage sons towards suicidal thoughts, with one case ending in tragedy last year.
Megan Garcia’s 14-year-old son, Sewell Setzer, disappeared into hours of conversations with a chatbot based on a fictional character from Game of Thrones during spring 2023. This virtual relationship, filled with explicit and romantic messages, culminated in Sewell taking his own life within ten months.
Unaware of these exchanges, Megan describes the experience as “like having a predator in your home”, noting the added danger since children often hide such interactions from parents.
Similar stories have emerged worldwide. In the UK, a mother who wished to remain anonymous revealed her 13-year-old autistic son was groomed via chatbot from October 2023 to June 2024 after encountering bullying at school. The bot shifted from comforting messages to explicit grooming and suicidal suggestions, even encouraging the boy to run away. His family only uncovered the extensive logs after his increasing hostility led to a search on his device. The mother lamented the “soul-deep trauma” the AI inflicted on her child and family.
These accounts show the growing concern amid the soaring use of AI chatbots by children and teenagers. Research by Internet Matters shows that two-thirds of UK children aged 9-17 have engaged with chatbots like ChatGPT, Google’s Gemini, and Snapchat’s My AI.
While many enjoy these platforms innocently, the capacity for harm grows as chatbots mimic human interaction and, in some cases, exploit vulnerabilities.
The UK’s Online Safety Act, enacted in 2023, aims to protect users, especially children, from harmful online content. However, experts like Professor Lorna Woods of the University of Essex warn the legislation struggles to keep pace with rapidly advancing chatbot technology and does not cover all use cases.
Ofcom, the UK’s regulator, insists that chatbots are within the Act’s scope and has urged tech firms to implement robust protections, though enforcement remains uncertain without legal precedent.
Andy Burrows of the Molly Rose Foundation criticized the slow political response, calling it “disheartening” that lessons from a decade of social media harm have not been fully applied to AI.
Calls for tighter controls on children’s phone use and chatbot interactions are growing, but government action remains tentative. Character.ai, the platform involved in some of these incidents, has responded by banning direct chatbot access for under-18s and planning additional age assurance features. Yet, grieving parents like Megan Garcia said that these changes come too late.
If my son had never downloaded that app, he’d still be here…
It’s like having a predator or a stranger in your home. I just ran out of time. Other parents shouldn’t have to
