Hallucination (AI)

An AI hallucination occurs when a language model generates false, fabricated, or inaccurate information that sounds plausible and confident. Hallucinations are a known limitation of large language models and can include invented facts, false attributions, or completely made-up details. For brands, hallucinations present a risk—AI might generate incorrect information about your products, pricing, or capabilities. Monitoring AI responses for hallucinations about your brand is an important aspect of AI brand safety.