DECEIVE-AFC: Adversarial Claim Attacks against Search-Enabled LLM-based Fact-Checking Systems

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ou, Haoran, Chen, Kangjie, Deng, Gelei, Liu, Hangcheng, Zhang, Jie, Zhang, Tianwei, Lam, Kwok-Yan
Format:	Preprint
Published:	2026
Subjects:	Cryptography and Security Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866917340169371648
author	Ou, Haoran Chen, Kangjie Deng, Gelei Liu, Hangcheng Zhang, Jie Zhang, Tianwei Lam, Kwok-Yan
author_facet	Ou, Haoran Chen, Kangjie Deng, Gelei Liu, Hangcheng Zhang, Jie Zhang, Tianwei Lam, Kwok-Yan
contents	Fact-checking systems with search-enabled large language models (LLMs) have shown strong potential for verifying claims by dynamically retrieving external evidence. However, the robustness of such systems against adversarial attack remains insufficiently understood. In this work, we study adversarial claim attacks against search-enabled LLM-based fact-checking systems under a realistic input-only threat model. We propose DECEIVE-AFC, an agent-based adversarial attack framework that integrates novel claim-level attack strategies and adversarial claim validity evaluation principles. DECEIVE-AFC systematically explores adversarial attack trajectories that disrupt search behavior, evidence retrieval, and LLM-based reasoning without relying on access to evidence sources or model internals. Extensive evaluations on benchmark datasets and real-world systems demonstrate that our attacks substantially degrade verification performance, reducing accuracy from 78.7% to 53.7%, and significantly outperform existing claim-based attack baselines with strong cross-system transferability.
format	Preprint
id	arxiv_https___arxiv_org_abs_2602_02569
institution	arXiv
publishDate	2026
record_format	arxiv
spellingShingle	DECEIVE-AFC: Adversarial Claim Attacks against Search-Enabled LLM-based Fact-Checking Systems Ou, Haoran Chen, Kangjie Deng, Gelei Liu, Hangcheng Zhang, Jie Zhang, Tianwei Lam, Kwok-Yan Cryptography and Security Artificial Intelligence Fact-checking systems with search-enabled large language models (LLMs) have shown strong potential for verifying claims by dynamically retrieving external evidence. However, the robustness of such systems against adversarial attack remains insufficiently understood. In this work, we study adversarial claim attacks against search-enabled LLM-based fact-checking systems under a realistic input-only threat model. We propose DECEIVE-AFC, an agent-based adversarial attack framework that integrates novel claim-level attack strategies and adversarial claim validity evaluation principles. DECEIVE-AFC systematically explores adversarial attack trajectories that disrupt search behavior, evidence retrieval, and LLM-based reasoning without relying on access to evidence sources or model internals. Extensive evaluations on benchmark datasets and real-world systems demonstrate that our attacks substantially degrade verification performance, reducing accuracy from 78.7% to 53.7%, and significantly outperform existing claim-based attack baselines with strong cross-system transferability.
title	DECEIVE-AFC: Adversarial Claim Attacks against Search-Enabled LLM-based Fact-Checking Systems
topic	Cryptography and Security Artificial Intelligence
url	https://arxiv.org/abs/2602.02569