Ministers are considering stronger restrictions on AI chatbots used by children, including mandatory breaks and action against services that present serious mental-health risks.
Why chatbots are different
A conversational system can appear patient, personal and authoritative. That makes it easy for a young user to overestimate what the system understands or to treat generated advice as confidential professional support.
Generative AI can produce inaccurate, inappropriate or overly agreeable responses. In a mental-health context, those failures may carry far greater consequences than an ordinary search error.
What safeguards should include
- Clear statements that the service is not a clinician.
- Reliable escalation to emergency and trusted-human support.
- Age-appropriate design and default usage breaks.
- Independent testing for harmful responses.
- Minimal collection and retention of sensitive conversations.
A role for schools and parents
Government plans also include updated online-safety guidance and stronger media literacy in RSHE. Young people need practical skills: checking claims, recognising manipulation, protecting personal information and knowing when a human professional is essential.
Regulation must avoid implying that a compliant chatbot is a safe substitute for care. The priority should be access to qualified support, with technology used only within clear boundaries.
