Hallucination

A model confidently generating information that sounds plausible but is fabricated or wrong. In agentic systems this is especially dangerous because the agent may act on false information.

Example: An agent invents a non-existent API endpoint and tries to call it, then fails to handle the 404 and spirals into a retry loop.

In practice: Grounding (via RAG or tool results) reduces hallucination by giving the model real data to work from. Verification steps — like having another agent fact-check — add another layer of defense.

• mitigated by RAG (Retrieval-Augmented Generation)Grounding the agent's responses in retrieved documents reduces the chance of the model fabricating information.