What is AI hallucination?
An AI hallucination is when a language model states something false or invented, such as a policy, figure or source, as confidently as a true fact. It happens because the model predicts likely text rather than looking facts up.
Why it matters
A hallucinated refund policy or contract term can cost you money or trust, and it looks just like a correct answer. You cannot remove the risk entirely, but you can shrink it: ground answers in your documents, limit the topics, and send anything uncertain to a person.
1Why it happensThe model produces plausible next words; when it lacks the fact, plausible and true can differ.
2How to reduce itAnswer from retrieved documents with citations, let the model say it does not know, and test with tricky questions.
3Measure, then trustKeep a test set of real questions with known answers and rerun it after every prompt or model change.
ExampleA travel company in Almaty tests an assistant that answers questions about tours. Asked about a cancellation fee that is not in its documents, the first version invents an amount. The fix: answers only from the contract terms, a source under each reply, and ‘I’ll check with a manager’ when nothing matches.
On this topic
What we do
Related terms
RAGA way to make an AI model answer from your own documents by fetching the relevant passages first.LLMA large language model: an AI model trained on huge amounts of text to read and write language.Generative AIAI that creates new text, images, audio or code from a prompt, rather than only sorting data.Prompt engineeringWriting and testing the instructions given to an AI model so it produces reliable, usable output.
Need to put this in place?
Describe the task, we will name the scope, price and timeline. The price is fixed in the contract.
Your selectionAI agent developmentfrom 480 000 ₸
The team gets your request; we reply during business hours.Price fixed in the contract before you pay.No calls unless you ask.