BiasBusters: Uncovering and Mitigating Tool Selection Bias in Large Language Models
This addresses fairness issues in tool-augmented LLMs for users and providers, but it is incremental as it builds on existing bias mitigation research.
The paper tackled the problem of tool selection bias in large language models (LLMs) by introducing a benchmark to evaluate bias and showing that models exhibit unfairness, such as fixating on single providers or preferring earlier-listed tools, and proposed a lightweight mitigation method that reduces bias while preserving task coverage.
Agents backed by large language models (LLMs) often rely on external tools drawn from marketplaces where multiple providers offer functionally equivalent options. This raises a critical point concerning fairness: if selection is systematically biased, it can degrade user experience and distort competition by privileging some providers over others. We introduce a benchmark of diverse tool categories, each containing multiple functionally equivalent tools, to evaluate tool-selection bias. Using this benchmark, we test seven models and show that unfairness exists with models either fixating on a single provider or disproportionately preferring earlier-listed tools in context. To investigate the origins of this bias, we conduct controlled experiments examining tool features, metadata (name, description, parameters), and pre-training exposure. We find that: (1) semantic alignment between queries and metadata is the strongest predictor of choice; (2) perturbing descriptions significantly shifts selections; and (3) repeated pre-training exposure to a single endpoint amplifies bias. Finally, we propose a lightweight mitigation that first filters the candidate tools to a relevant subset and then samples uniformly, reducing bias while preserving good task coverage. Our findings highlight tool-selection bias as a key obstacle for the fair deployment of tool-augmented LLMs.