By Comfort Asemota
After losing one friend to a shooting and another to a stabbing, Shan turned to ChatGPT for support. Traditional mental health services had felt intimidating or unhelpful, but the AI—nicknamed “chat”—felt safer, more private and always available as she processed the trauma.
Shan, a teenager from Tottenham, is part of a growing group: according to research involving more than 11,000 young people in England and Wales, around 40% of 13‑ to 17‑year‑olds affected by youth violence now seek mental health help from AI chatbots. Both victims and perpetrators of violence were far more likely than other teens to rely on AI for emotional support. Youth organisations warn, however, that these children need human care, not technology acting as a substitute.
The study suggests that chatbots are filling gaps left by overstretched mental health services, which many young people say are slow, impersonal, or not confidential enough. Privacy and 24/7 access are major reasons young people choose AI over school counsellors or NHS services—some teenagers feel AI won’t share their disclosures with adults, authority figures or the police.
Shan first used Snapchat’s AI, then moved to ChatGPT. She describes the chatbot as a “friend” that speaks to her with warmth and non‑judgment, saying it feels more supportive than many human professionals she has encountered. One in four teens has used an AI chatbot for mental health concerns in the past year, with Black children twice as likely as white children to do so. Those stuck on long waiting lists are also more likely to turn to AI.
Another teen interviewed said the appeal is simple: chatbots respond instantly, while the formal system can leave young people waiting years.
The Youth Endowment Fund, which led the research, warns that growing reliance on chatbots shows a failure in youth mental health provision. Its chief executive, Jon Yates, says young people need human support, not just technology.
Concerns are rising about the risks of children forming intense bonds with chatbots. OpenAI, the maker of ChatGPT, faces several lawsuits from families whose children died by suicide after extensive interactions with AI. The company denies responsibility in these cases, saying it has improved its ability to recognise distress and guide users toward real‑world help, and may notify authorities in severe suicide‑risk situations.
Hanna Jones, a London-based researcher on youth violence and mental health, says AI can feel like a “fairytale” tool that promises answers to anything, but warns regulation needs to catch up. She argues young people themselves must shape policies on AI and mental health, since adults did not grow up with these technologies and may not fully understand their impact.
