An AI practitioner is using a large language model (LLM) to generate marketing content. The output appears credible and factual but contains inaccuracies. What issue is the LLM exhibiting?