🌀 AI Hallucination
4th Grade 🤖 Artificial Intelligence
A hallucination is when an AI tool makes up information that sounds believable but isn’t true — a fact that doesn’t exist, a quote nobody said or a book that was never written.
A language model builds its answer one word at a time based on what seems likely. Sometimes the most “likely” thing, as far as the pattern goes, isn’t true at all in real life. For example, if you ask about a book that doesn’t exist, the model might invent a plot and an author, because that looks like a normal answer about a book.
Hallucinations are extra tricky because they’re written with the same confidence and style as correct answers. There’s no warning sign that says “I made this part up.” So for anything important — schoolwork, health, numbers or dates — check a reliable source, like a book, an official website or a grown-up who knows the subject.
A common mistake is thinking a hallucination is a lie on purpose. AI has no intentions; it doesn’t “know” it’s wrong. The hallucination comes from the way the model works — guessing from patterns. And people make mistakes too, so checking every source is a good habit.
For example
- A chatbot that invents a famous quote and says a person said it when they never did.
- An answer with a list of sources — but some of the books on the list don’t exist.
- A model that gives the wrong date for a historical event with total confidence.
- A model that makes up details about a little-known person because it “completed the pattern.”
📌 Remember: Hallucination = a made-up answer that sounds sure. Sounds good ≠ true. Check it!
Quick quiz
What is an AI hallucination?
Why is a hallucination hard to spot?
How can you protect yourself from hallucinations?