A hallucination is when an AI model produces information that sounds plausible and confident but is factually wrong or made up. It happens because language models predict likely text rather than checking a source of truth. For example, a model might invent a citation or a statistic that does not exist. This is why AI output on important matters needs to be verified.
Want this applied in your business? See how we take it to production:
We build this AI in production — at fixed prices, with one named expert. Start with a free consultation.
Book a free consultation →