This article describes how to completely avoid AI hallucinations with ChatGPT by adding an instruction that tells the AI not to answer unless the answer is within the specified context.