Why Do AI Agents Hallucinate?
Agent loops amplify LLM hallucinations by compounding errors across downstream steps
“one hallucinated fact in step two can poison step three, step four and everything downstream”
A LangChain explainer argues that while LLMs hallucinate at the model level, agentic loops make this more dangerous by compounding errors across multi-step pipelines. The recommended mitigation is evals rather than informal vibe-checking. This is a useful framing for practitioners building agent systems but covers well-established ground.