LLM Agents Can Be Choice-Supportive Biased Evaluators: An Empirical Study

With Large Language Model (LLM) agents taking on more evaluation responsibilities in decision-making, it is essential to recognize their possible biases to guarantee fair and trustworthy AI-supported decisions. This study is the first to thoroughly examine the choice-supportive bias in LLM agents, a cognitive bias that is known to impact human decisionmaking and evaluation. We conduct experiments across 19 open/unopen-source LLM models in five scenarios at maximum, employing both memory-based and evaluation-based tasks adapted and redesigned from human cognitive studies. Our findings show that LLM agents may exhibit biased attribution or evaluation that supports their initial choices, and such bias may persist even if contextual hallucination is not observable. Key findings show that bias manifestation can differ greatly depending on prompt construction and context preservation, and the bias may be mitigated in larger models. Significantly, we observe that the bias increases when the agents perceive they are in control. Our extensive study involving 284 well-educated humans shows that, despite bias, certain LLM agents can still perform better than humans in similar evaluation tasks. This research contributes to the growing area of AI psychology, and the findings underscore the importance of addressing cognitive biases in LLM Agent systems, with wide-ranging implications spanning from improving AI-assisted decision-making to advancing AI safety and ethics. * These authors contributed equally. All equal-contributing authors worked jointly from the first day until the last day, sharing equal workload and contributions. N. Zhuang, B. Cao, M. Xu were responsible for designing and conducting the experiments, Y. Yang was responsible for coding, J. Xu performed the human study, Y. Wang and Q. Liu were responsible for reviewing and supervising.