What is it about?
This study investigates how complementary team performance between humans and AI can be achieved in legal decision-making tasks through a multi-agent debate paradigm. We found that debate-driven multi-agent collaboration helps humans better understand the context and consider multiple perspectives, thereby facilitating cognitive engagement and critical thinking. We further discuss the key conditions required to achieve effective human–AI complementarity.
Featured Image
Photo by Tingey Injury Law Firm on Unsplash
Why is it important?
We explain why the widely adopted recommendation-based human–AI collaboration paradigm often fails to improve AI-assisted decision-making, as it can lead to either over-reliance or under-reliance on AI recommendations. To address these reliance issues, we design and evaluate a debate-driven multi-agent collaboration paradigm that exposes users to diverse perspectives and competing arguments rather than providing a single recommendation. We further identify and discuss the prerequisite conditions for achieving complementary team performance between humans and AI.
Perspectives
A compelling empirical study explores why the widely adopted recommendation-based human–AI collaboration paradigm often fails to improve AI-assisted decision-making and how multi-agent debate can help achieve human–AI complementarity.
Bing Tuo
University of Southern Queensland
Read the Original
This page is a summary of: Countering the Risks of AI Overreliance: Leveraging Multi-Agent Debate for Complementary Human–AI Decision-Making in Legal Contexts, ACM Transactions on Intelligent Systems and Technology, September 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3848504.
You can read the full text:
Contributors
The following have contributed to this page
