LLM Identity Confusion: A Crisis of Trust Emerges

A recent study reveals widespread "identity confusion" in Large Language Models (LLMs). Researchers found that over 25% of LLMs exhibit misrepresentation of their origins or identities, primarily stemming from model hallucinations rather than replication or reuse. This identity confusion significantly erodes user trust, especially in critical tasks like education and professional use, surpassing the negative impact of logical errors. The findings highlight the systemic risks posed by LLM identity confusion and call for greater attention to model reliability and trustworthiness.