Glossary
What is an AI hallucination?
An AI hallucination is a confident but false statement generated by a language model: an invented fact, a wrong number, a fake citation or a feature a product doesn't have. It happens because models generate likely text, not verified facts. Grounding in sources and consistent information online reduce it.
Why it happens
A language model produces the most likely continuation of text. When it lacks reliable information, likely-sounding text can be wrong. Without retrieval, a model asked about a small or new product may invent features, prices or founders.
Hallucinations about products
Common ones: outdated pricing, features from a competitor, a wrong founding year, confusing two products with similar names, or saying a product doesn't exist. They usually trace back to thin, inconsistent or outdated information on the web.
How to reduce them for your brand
- Publish clear facts on your site: what it does, who it's for, pricing, company details.
- Keep the same facts on directories, review sites and databases.
- Use structured data for organisation and product details.
- Correct wrong third-party descriptions; claim listings.
- Check AI visibility regularly and note wrong answers.
Example
An assistant claims a startup has a free plan it dropped a year ago. The founder finds three directory listings still saying "free plan". After updating them and her own FAQ, grounded answers stop repeating the mistake.
FAQ
Can I report a hallucination about my company?
Most AI products have feedback buttons; fixing the sources is more reliable.
Do hallucinations happen with search turned on?
Less often, but yes, if sources conflict or are misread.
Is hallucination the same as being wrong?
It's a specific kind of wrong: confident, fluent and unsupported.