What is hallucination?
Quick answer
A hallucination is when a generative AI states something false as though it were fact, usually because the model is predicting plausible language rather than retrieving verified information.
Every AI-assisted assignment carries a verification log: claim, source checked, verdict. It takes five minutes and changes how students read output for good.
In context
Hallucinations are not bugs to be patched away. They follow from how these systems work, which is why verification is a permanent skill rather than a temporary workaround.
Students who understand this stop treating an assistant as an oracle and start treating it as a fast, unreliable first draft.
In a NASCA classroom
Every AI-assisted assignment carries a verification log: claim, source checked, verdict. It takes five minutes and changes how students read output for good.
Related terms
Generative AI
Generative AI produces new text, images, audio or code by predicting what should come next, based on patterns learned from very large amounts of training data.
AI literacy
AI literacy is the ability to recognise where AI is being used, understand roughly how it works, judge its output, and decide when using it is appropriate.
Prompt engineering
Prompt engineering is the practice of writing instructions that get a reliable, specific result from an AI model, using context, constraints, examples and a clear description of the output you want.