Examples of AI Hallucinations

Last updated June 2026

Hallucinations are easier to understand with examples. Here are common patterns drawn from how models fail in the real world.

Fake legal citations

In well-documented incidents, lawyers have submitted AI-generated briefs citing court cases that never existed. The model produced realistic case names, numbers, and quotes — all fabricated. This is the canonical example of why you must verify every citation.

Don't just trust — verify

Run your question through ChatVerify and compare answers across leading AI systems.

Check AI Consensus

Invented sources and studies

Ask a model for studies supporting a claim and it may produce authoritative-sounding references — authors, journals, years — that don't exist. The format is convincing; the content is fiction.

Wrong numbers and outdated facts

Models routinely state incorrect statistics, prices, and dates with full confidence, especially for fast-changing topics. Without a source, treat any number as unverified.

Related reading

Verify before you act

AI gives answers. ChatVerify helps you verify them.