@inproceedings{zhao-etal-2026-mark,
title = "The Mark Fades: Adaptive Evolutionary Paraphrase-based Attack against {LLM} Watermarks",
author = "Zhao, Yusheng and
Zhao, Jian and
Zhang, Tianle and
Wei, Feng and
Li, Xuelong",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.459/",
pages = "9428--9445",
ISBN = "979-8-89176-395-1",
abstract = "While LLM watermarking is essential for machine- generated content identification, existing paraphrase-based attacks struggle to balance watermark removal efficacy with text quality. We propose TSAPA, a training-free evolutionary framework that models watermark removal as a constrained multi-objective optimization problem. By leveraging genetic algorithms to navigate the Pareto front, TSAPA utilizes a Pseudo-Log-Likelihood (PLL)-guided mutation to precisely target and modify watermark-carrying tokens. Experiments on Qwen3 series (1.7B/8B/32B) across multiple watermark schemes show that TSAPA achieves over 90{\%} attack success rate (ASR) while maintaining high text semantic fidelity, significantly outperforming baselines methods. This work exposes critical vulnerabilities in current watermarks and provides a new perspective for robust evaluation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhao-etal-2026-mark">
<titleInfo>
<title>The Mark Fades: Adaptive Evolutionary Paraphrase-based Attack against LLM Watermarks</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yusheng</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jian</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianle</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Feng</namePart>
<namePart type="family">Wei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuelong</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>While LLM watermarking is essential for machine- generated content identification, existing paraphrase-based attacks struggle to balance watermark removal efficacy with text quality. We propose TSAPA, a training-free evolutionary framework that models watermark removal as a constrained multi-objective optimization problem. By leveraging genetic algorithms to navigate the Pareto front, TSAPA utilizes a Pseudo-Log-Likelihood (PLL)-guided mutation to precisely target and modify watermark-carrying tokens. Experiments on Qwen3 series (1.7B/8B/32B) across multiple watermark schemes show that TSAPA achieves over 90% attack success rate (ASR) while maintaining high text semantic fidelity, significantly outperforming baselines methods. This work exposes critical vulnerabilities in current watermarks and provides a new perspective for robust evaluation.</abstract>
<identifier type="citekey">zhao-etal-2026-mark</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.459/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>9428</start>
<end>9445</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The Mark Fades: Adaptive Evolutionary Paraphrase-based Attack against LLM Watermarks
%A Zhao, Yusheng
%A Zhao, Jian
%A Zhang, Tianle
%A Wei, Feng
%A Li, Xuelong
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F zhao-etal-2026-mark
%X While LLM watermarking is essential for machine- generated content identification, existing paraphrase-based attacks struggle to balance watermark removal efficacy with text quality. We propose TSAPA, a training-free evolutionary framework that models watermark removal as a constrained multi-objective optimization problem. By leveraging genetic algorithms to navigate the Pareto front, TSAPA utilizes a Pseudo-Log-Likelihood (PLL)-guided mutation to precisely target and modify watermark-carrying tokens. Experiments on Qwen3 series (1.7B/8B/32B) across multiple watermark schemes show that TSAPA achieves over 90% attack success rate (ASR) while maintaining high text semantic fidelity, significantly outperforming baselines methods. This work exposes critical vulnerabilities in current watermarks and provides a new perspective for robust evaluation.
%U https://aclanthology.org/2026.findings-acl.459/
%P 9428-9445
Markdown (Informal)
[The Mark Fades: Adaptive Evolutionary Paraphrase-based Attack against LLM Watermarks](https://aclanthology.org/2026.findings-acl.459/) (Zhao et al., Findings 2026)
ACL