Chain of Attack: On the Robustness of Vision-Language Models Against Transfer-Based Adversarial Attacks
This work addresses safety concerns for users of vision-language models by exposing vulnerabilities through a novel attack method, though it is incremental as it builds on existing transfer-based strategies.
The paper tackles the problem of evaluating the robustness of vision-language models against transfer-based adversarial attacks by proposing Chain of Attack (CoA), which iteratively enhances adversarial example generation using multi-modal semantic updates, achieving superior transferability and efficiency in misleading models to generate targeted responses in black-box scenarios.
Pre-trained vision-language models (VLMs) have showcased remarkable performance in image and natural language understanding, such as image captioning and response generation. As the practical applications of vision-language models become increasingly widespread, their potential safety and robustness issues raise concerns that adversaries may evade the system and cause these models to generate toxic content through malicious attacks. Therefore, evaluating the robustness of open-source VLMs against adversarial attacks has garnered growing attention, with transfer-based attacks as a representative black-box attacking strategy. However, most existing transfer-based attacks neglect the importance of the semantic correlations between vision and text modalities, leading to sub-optimal adversarial example generation and attack performance. To address this issue, we present Chain of Attack (CoA), which iteratively enhances the generation of adversarial examples based on the multi-modal semantic update using a series of intermediate attacking steps, achieving superior adversarial transferability and efficiency. A unified attack success rate computing method is further proposed for automatic evasion evaluation. Extensive experiments conducted under the most realistic and high-stakes scenario, demonstrate that our attacking strategy can effectively mislead models to generate targeted responses using only black-box attacks without any knowledge of the victim models. The comprehensive robustness evaluation in our paper provides insight into the vulnerabilities of VLMs and offers a reference for the safety considerations of future model developments.