cs.AINov 17, 2026

Not All Evidence Is Admissible: Evidence Governance for Tool-Augmented Medical Agents

Authors: Runxun Zhang, Onofrio Antonio Catalano, Jinqiao Wang, Jingwei Wei

Organizations: Institute of Automation, Chinese Academy of Sciences · Massachusetts General Hospital · University of Chinese Academy of Sciences

Abstract

Tool-augmented medical agents treat whatever their tools return as evidence. We show that this default is costly: medical tools fail silently, and a frozen reasoner cannot ignore what enters its context, so many cases it would have answered correctly are flipped to wrong once tool outputs are added. We call this evidence hijacking. V\mathcal{V}-information explains the phenomenon: the reasoner has lost the option to ignore, so a tool output can carry negative usable information, and moving the decision of what to keep outside the reasoner restores that option. The theory further shows that part of any filtering gain comes merely from letting the reasoner see less, and only selection by content lets it see better. Building on this, we recast tool-augmented reasoning as evidence governance, which separates deciding what counts as evidence from the reasoner that answers, and propose context-independent evidence governance (CIEG), the first medical agent framework that rules, after acquisition, on whether each individual tool output may enter as evidence. Every tool output is written as a card that can be checked against the raw input; an independent checker that sees only the input and the cards decides which cards become evidence; and the reasoner answers from the admitted cards alone, with no training anywhere. On medical question answering, fundus glaucoma screening, and chest X-ray pneumonia classification, it achieves the highest accuracy among models of comparable scale, and significantly outperforms state-of-the-art tool-using agents and multi-agent systems built on the same backbone at lower cost.

Figures & tables

Explore similar work

CardsList
  1. From Given to Gathered Evidence: Agentic Learning for Longitudinal Medical Reasoning

    Sep 30, 2026Minye Shao, Chaohui Yu, Yixuan Wu +6AI Agent BenchmarksMedical VLMs

  2. X-Ragent: Tool-Augmented Reinforcement Learning for Evidence-Grounded Chest X-Ray Reasoning

    Oct 12, 2026Yi Wang, Zhenlong Yuan, Yuchan Guo +9Medical VLMsMedical Imaging

  3. Mind the Tool Failures: Achieving Synergistic Tool Gains for Medical Agents

    May 26, 2026Yunhui Gan, Tan Pan, Kaiyu Guo +5Tool UseRL for Tool Use