Textual backdoor attacks pose a practical threat to existing systems, as they can compromise the model by inserting imperceptible triggers into inputs and manipulating labels in the training dataset. With cutting-edge generative models such as GPT-4 pushing rewriting to extraordinary levels, such attacks are becoming even harder to detect. We conduct a comprehensive investigation of the role of black-box generative models as a backdoor attack tool, highlighting the importance of researching relative defense strategies. In this paper, we reveal that the proposed generative model-based attack, BGMAttack, could effectively deceive textual classifiers. Compared with the traditional attack methods, BGMAttack makes the backdoor trigger less conspicuous by leveraging state-of-the-art generative models. Our extensive evaluation of attack effectiveness across five datasets, complemented by three distinct human cognition assessments, reveals that Figure 4 achieves comparable attack performance while maintaining superior stealthiness relative to baseline methods.
@article{arxiv.2304.14475,
title = {ChatGPT as an Attack Tool: Stealthy Textual Backdoor Attack via Blackbox Generative Model Trigger},
author = {Jiazhao Li and Yijin Yang and Zhuofeng Wu and V. G. Vinod Vydiswaran and Chaowei Xiao},
journal= {arXiv preprint arXiv:2304.14475},
year = {2023}
}