Hallucination (AI)
When an AI chatbot generates information that sounds confident and plausible but is entirely fabricated. AI tools do not verify facts, they predict likely-sounding words.
Hallucination (AI) is when an AI chatbot generates information that sounds confident and plausible but is entirely fabricated, because the tool predicts likely-sounding words rather than verifying facts.
In plain English
When an AI chatbot generates information that sounds confident and plausible but is entirely fabricated. AI tools do not verify facts, they predict likely-sounding words.
Why it matters
Children using AI for homework, mental-health questions, or safety advice can be given confidently wrong answers. Teaching them to check claims against trusted sources, and to never rely on a chatbot for medical, legal, or safeguarding advice, is essential.
Related
Risks and topics
Frequently Asked Questions
This is practical educational content to support families. For case-specific concerns about a child's safety, contact the NSPCC helpline on 0808 800 5000 or your local safeguarding team.
Last reviewed: 2026-05-17