Chair personas affect timing and bias in AI clinical decisions
Consistent but Biased: Chair Personas Trade Noise for Bias in Multi-Agent Clinical Decisions
Multiagent Systems
Summary
The authors studied how different personalities assigned to a lead AI agent influenced medical decisions in a team of AI doctors. They found that changing the chair agent's persona altered when decisions were made, trading careful timing for bias in those decisions. Some personas made decisions more quickly or with delays, but didn’t strongly improve agreement or reliability between repeated AI runs. This shows that AI systems should report how the lead agent is prompted, as it affects outcomes beyond just the model used.
What this means in practice
- •For clinical ai developers: Design and test persona prompts for lead AI agents to control biases in timing and decision style in medical AI teams.
- •For software engineers building multi-agent systems: Adjust persona prompts on chair agents to influence group decision outcomes and dynamics in AI collaborative systems.
Authors
Yitong Bai, Chengyi Zhang, Ziyang Wang
Abstract
In clinical systems built from several large language model (LLM) agents, one agent often chairs the discussion and makes the final decision. Whether the chair's persona prompt changes that decision, or the agreement between repeated runs that some systems treat as reliability, has not been tested with the case held constant. We ask whether the persona changes the decision or only its timing, how it affects agreement between runs, and which behaviors appear when the prompt describes only personality traits. A simulated department of seven agents discussed 20 published cases of degenerative cervical myelopathy, with the case reports replayed unchanged in every run. Only the prompts of the chair and its deputy changed (no persona, or a consultative, autocratic or avoidant one), written either as rules of conduct or as personality traits. With personas written as rules, the chair chose to operate now in 93% of runs when autocratic and in 2% when avoidant, against 57% without a persona. Surgery was still planned in nearly all decisions, so the personas mainly changed whether to operate now or after more tests. These effects held with a second model and reworded prompts. The autocratic and avoidant personas made the three runs of a case agree more often, but only about as often as chance predicts, and Fleiss $κ$ did not rise. Prompts that described only personality traits kept more than half of the autocratic and avoidant effects on operating now. They reproduced quick decisions, delays and extra tests, but not ignoring red flags, dropping backup plans or passing the decision to the family. Studies should report and vary the chair's persona as they do the model, and correct agreement between runs for chance before calling it reliability.