BadApex: Backdoor Attack Based on Adaptive Optimization Mechanism of Black-box Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wu, Zhengxian, Wen, Juan, Peng, Wanli, Zhang, Ziwei, Zhou, Yinghan, Xue, Yiming
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916698244775936
author Wu, Zhengxian
Wen, Juan
Peng, Wanli
Zhang, Ziwei
Zhou, Yinghan
Xue, Yiming
author_facet Wu, Zhengxian
Wen, Juan
Peng, Wanli
Zhang, Ziwei
Zhou, Yinghan
Xue, Yiming
contents Previous insertion-based and paraphrase-based backdoors have achieved great success in attack efficacy, but they ignore the text quality and semantic consistency between poisoned and clean texts. Although recent studies introduce LLMs to generate poisoned texts and improve the stealthiness, semantic consistency, and text quality, their hand-crafted prompts rely on expert experiences, facing significant challenges in prompt adaptability and attack performance after defenses. In this paper, we propose a novel backdoor attack based on adaptive optimization mechanism of black-box large language models (BadApex), which leverages a black-box LLM to generate poisoned text through a refined prompt. Specifically, an Adaptive Optimization Mechanism is designed to refine an initial prompt iteratively using the generation and modification agents. The generation agent generates the poisoned text based on the initial prompt. Then the modification agent evaluates the quality of the poisoned text and refines a new prompt. After several iterations of the above process, the refined prompt is used to generate poisoned texts through LLMs. We conduct extensive experiments on three dataset with six backdoor attacks and two defenses. Extensive experimental results demonstrate that BadApex significantly outperforms state-of-the-art attacks. It improves prompt adaptability, semantic consistency, and text quality. Furthermore, when two defense methods are applied, the average attack success rate (ASR) still up to 96.75%.
format Preprint
id arxiv_https___arxiv_org_abs_2504_13775
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BadApex: Backdoor Attack Based on Adaptive Optimization Mechanism of Black-box Large Language Models
Wu, Zhengxian
Wen, Juan
Peng, Wanli
Zhang, Ziwei
Zhou, Yinghan
Xue, Yiming
Computation and Language
Cryptography and Security
Previous insertion-based and paraphrase-based backdoors have achieved great success in attack efficacy, but they ignore the text quality and semantic consistency between poisoned and clean texts. Although recent studies introduce LLMs to generate poisoned texts and improve the stealthiness, semantic consistency, and text quality, their hand-crafted prompts rely on expert experiences, facing significant challenges in prompt adaptability and attack performance after defenses. In this paper, we propose a novel backdoor attack based on adaptive optimization mechanism of black-box large language models (BadApex), which leverages a black-box LLM to generate poisoned text through a refined prompt. Specifically, an Adaptive Optimization Mechanism is designed to refine an initial prompt iteratively using the generation and modification agents. The generation agent generates the poisoned text based on the initial prompt. Then the modification agent evaluates the quality of the poisoned text and refines a new prompt. After several iterations of the above process, the refined prompt is used to generate poisoned texts through LLMs. We conduct extensive experiments on three dataset with six backdoor attacks and two defenses. Extensive experimental results demonstrate that BadApex significantly outperforms state-of-the-art attacks. It improves prompt adaptability, semantic consistency, and text quality. Furthermore, when two defense methods are applied, the average attack success rate (ASR) still up to 96.75%.
title BadApex: Backdoor Attack Based on Adaptive Optimization Mechanism of Black-box Large Language Models
topic Computation and Language
Cryptography and Security
url https://arxiv.org/abs/2504.13775