What is AI Hallucination?
Last updated May 2026When an AI model generates plausible-sounding but factually incorrect information.
Definition
An AI hallucination occurs when a language model generates information that sounds convincing and fluent but is factually wrong, fabricated, or unsupported. Hallucinations happen because LLMs predict statistically likely text rather than retrieving verified facts. They can invent citations, make up statistics, or confidently state incorrect information.
💡 Example
If you ask an AI about a specific court case and it generates a detailed summary with a case number that does not exist, that is a hallucination. The AI produced text that followed the pattern of legal writing but contained fabricated details.
Related concepts
A type of AI trained on massive text datasets to understand and generate human language.
A technique that lets AI access external knowledge bases to provide more accurate answers.
Why this matters
AI hallucination — when models generate plausible-sounding but false information — is the biggest trust barrier in AI adoption. Understanding it helps you design better prompts, choose tools with built-in fact-checking, and know when to verify AI outputs.
Real-world example
Ask ChatGPT for a citation and it might invent a paper that doesn't exist. Ask Perplexity the same question and it cites real sources you can verify. This is why search-grounded tools reduce hallucination risk compared to pure language models.
See it in action
Connecting AI outputs to verified sources of information to reduce hallucinations.
Explore AI tools
Find tools that use ai hallucination in practice.
What is AI Hallucination?
An AI hallucination occurs when a language model generates information that sounds convincing and fluent but is factually wrong, fabricated, or unsupported. Hallucinations happen because LLMs predict statistically likely text rather than retrieving verified facts. They can invent citations, make up statistics, or confidently state incorrect information.
How does AI Hallucination work in practice?
If you ask an AI about a specific court case and it generates a detailed summary with a case number that does not exist, that is a hallucination. The AI produced text that followed the pattern of legal writing but contained fabricated details.
How can you detect AI hallucinations?
Check factual claims against reliable sources, look for overly specific details (exact dates, statistics) that seem too precise, and watch for confident language about obscure topics. Cross-referencing AI output with a second model or search engine helps catch fabricated information.
Which AI tools hallucinate the least?
RAG-based tools like Perplexity and NotebookLM tend to hallucinate less because they ground responses in retrieved sources. Among general chatbots, newer models with better training generally hallucinate less, but no model is hallucination-free. Always verify important claims.
Why do AI models hallucinate?
Models hallucinate because they generate text based on statistical patterns rather than factual understanding. They are trained to produce fluent, plausible-sounding text, which means they can confidently state things that sound right but are fabricated. Gaps in training data increase hallucination risk.