Skip to content
Sitebard AI
Generative AI

Hallucination

When an AI model produces output that sounds confident and plausible but is factually incorrect or fabricated.

By Sitebard TeamUpdated April 3, 2026

In plain English

An AI hallucination is when a chatbot confidently says something that is simply not true. It can invent facts, names, or sources that look real but are made up.

Technical definition

A hallucination is an output from a generative model that is unfaithful to its training data, provided context, or factual reality. It arises because such models optimize for statistically likely continuations rather than verified truth, and is often mitigated with retrieval grounding and constrained decoding.

Business use case

Understanding hallucination helps businesses design safe AI workflows by adding source grounding, citations, and human review where accuracy is critical. This protects against publishing or acting on fabricated information in customer-facing or regulated contexts.

Example

Asked for a citation, a chatbot invents a realistic-looking research paper title, author, and journal that does not actually exist.

Frequently asked questions

An AI hallucination is when a model generates information that is false, fabricated, or not supported by its sources, yet presents it in a confident and convincing way.

Language models predict likely text rather than verify facts, so when they lack reliable information they may fill gaps with plausible-sounding but incorrect content.

Techniques include grounding responses in retrieved sources, asking the model to cite evidence, lowering creativity settings, and adding human review for high-stakes outputs.

Keep exploring

View all

Put AI intelligence to work in your business

Sitebard AI brings together the data, guides, and career intelligence you need to make confident AI decisions.