AI hallucinations are confident but false outputs that pose major security risks. Learn how they impact threat detection and how to mitigate them.