# Marking the Wrong Symptoms: Evaluating LLM Watermarks in Medical Texts

arXiv: https://arxiv.org/abs/2607.20462 · HTML: https://mlsanity.com/paper/2607.20462/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Watermarking generated content and models, one of 12 directions in LLM safety and security.
- Builds on 23 of its 40 references that are in the map.
- First posted on arXiv 16 May 2026.

## Placed in

- LLM safety and security / [Watermarking generated content and models](https://mlsanity.com/safety/watermarking-generated-content-and-models/)

## Builds on (most cited first, 23 of 23)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Gemma 3 Technical Report](https://mlsanity.com/paper/2503.19786/) (2025)
- [What Disease does this Patient Have? A Large-scale Open Domain Question Answering Dataset from Medical Exams](https://mlsanity.com/paper/2009.13081/) (2020)
- [A Watermark for Large Language Models](https://mlsanity.com/paper/2301.10226/) (2023)
- [Phi-4 Technical Report](https://mlsanity.com/paper/2412.08905/) (2024)
- [MedGemma Technical Report](https://mlsanity.com/paper/2507.05201/) (2025)
- [HealthBench: Evaluating Large Language Models Towards Improved Human Health](https://mlsanity.com/paper/2505.08775/) (2025)
- [Robust Distortion-free Watermarks for Language Models](https://mlsanity.com/paper/2307.15593/) (2023)
- [Lingshu: A Generalist Foundation Model for Unified Multimodal Medical Understanding and Reasoning](https://mlsanity.com/paper/2506.07044/) (2025)
- [MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding](https://mlsanity.com/paper/2501.18362/) (2025)
- [OmniMedVQA: A New Large-Scale Comprehensive Evaluation Benchmark for Medical LVLM](https://mlsanity.com/paper/2402.09181/) (2024)
- [Med-HALT: Medical Domain Hallucination Test for Large Language Models](https://mlsanity.com/paper/2307.15343/) (2023)
- [Who Wrote this Code? Watermarking for Code Generation](https://mlsanity.com/paper/2305.15060/) (2023)
- [A Survey of Text Watermarking in the Era of Large Language Models](https://mlsanity.com/paper/2312.07913/) (2023)
- [GMAI-MMBench: A Comprehensive Multimodal Evaluation Benchmark Towards General Medical AI](https://mlsanity.com/paper/2408.03361/) (2024)
- [UltraMedical: Building Specialized Generalists in Biomedicine](https://mlsanity.com/paper/2406.03949/) (2024)
- [WaterBench: Towards Holistic Evaluation of Watermarks for Large Language Models](https://mlsanity.com/paper/2311.07138/) (2023)
- [WatME: Towards Lossless Watermarking Through Lexical Redundancy](https://mlsanity.com/paper/2311.09832/) (2023)
- [Evaluating Reasoning Faithfulness in Medical Vision-Language Models using Multimodal Perturbations](https://mlsanity.com/paper/2510.11196/) (2025)
- [Factuality Beyond Coherence: Evaluating LLM Watermarking Methods for Medical Texts](https://mlsanity.com/paper/2509.07755/) (2025)
- [A Unified Framework for LLM Watermarks](https://mlsanity.com/paper/2602.06754/) (2026)
- [Right Prediction, Wrong Reasoning: Uncovering LLM Misalignment in RA Disease Diagnosis](https://mlsanity.com/paper/2504.06581/) (2025)
- [Enhancing Watermarking Quality for LLMs via Contextual Generation States Awareness](https://mlsanity.com/paper/2506.07403/) (2025)

## Abstract (arXiv)

Large language models (LLMs) are increasingly integrated into clinical workflows, stressing the need for reliable traceability of model-generated output with watermarking. Yet, most watermarks are evaluated on general-purpose benchmarks, leaving domains like medicine, where small token-level perturbations can result in significant semantic changes, underexplored. In this work, we present the first rigorous study of how LLM watermarks affect medical performance, benchmarking 5 watermarking schemes across 11 LLMs and 7 VLMs on various tasks spanning unimodal and multimodal clinical reasoning. Importantly, we complement existing evaluations by introducing a human-expert-validated pipeline for systematically auditing medical reasoning quality, terminological precision, and induced hallucinations. Our results reveal that watermarking can induce substantial degradation across multiple failure modes, including lexical corruption, hallucinated terminology, and amplified misattribution or omission of image findings. Notably, we find that the absence of domain-specific analyses, combined with aggregate metrics that miss failures inherent to clinical text, can systematically obscure practical watermark-induced degradations. Our findings establish domain-specific evaluation as a prerequisite for the safe deployment of watermarked models in medicine, where current benchmarks can otherwise mask clinically consequential failures.