CAHS-Attack: CLIP-Aware Heuristic Search Attack Method for Stable Diffusion
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915638898851840 |
|---|---|
| author | Xia, Shuhan Dai, Jing Ouyang, Hui Shang, Yadong Zhao, Dongxiao Li, Peipei |
| author_facet | Xia, Shuhan Dai, Jing Ouyang, Hui Shang, Yadong Zhao, Dongxiao Li, Peipei |
| contents | Diffusion models exhibit notable fragility when faced with adversarial prompts, and strengthening attack capabilities is crucial for uncovering such vulnerabilities and building more robust generative systems. Existing works often rely on white-box access to model gradients or hand-crafted prompt engineering, which is infeasible in real-world deployments due to restricted access or poor attack effect. In this paper, we propose CAHS-Attack , a CLIP-Aware Heuristic Search attack method. CAHS-Attack integrates Monte Carlo Tree Search (MCTS) to perform fine-grained suffix optimization, leveraging a constrained genetic algorithm to preselect high-potential adversarial prompts as root nodes, and retaining the most semantically disruptive outcome at each simulation rollout for efficient local search. Extensive experiments demonstrate that our method achieves state-of-the-art attack performance across both short and long prompts of varying semantics. Furthermore, we find that the fragility of SD models can be attributed to the inherent vulnerability of their CLIP-based text encoders, suggesting a fundamental security risk in current text-to-image pipelines. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_21180 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CAHS-Attack: CLIP-Aware Heuristic Search Attack Method for Stable Diffusion Xia, Shuhan Dai, Jing Ouyang, Hui Shang, Yadong Zhao, Dongxiao Li, Peipei Cryptography and Security Artificial Intelligence Diffusion models exhibit notable fragility when faced with adversarial prompts, and strengthening attack capabilities is crucial for uncovering such vulnerabilities and building more robust generative systems. Existing works often rely on white-box access to model gradients or hand-crafted prompt engineering, which is infeasible in real-world deployments due to restricted access or poor attack effect. In this paper, we propose CAHS-Attack , a CLIP-Aware Heuristic Search attack method. CAHS-Attack integrates Monte Carlo Tree Search (MCTS) to perform fine-grained suffix optimization, leveraging a constrained genetic algorithm to preselect high-potential adversarial prompts as root nodes, and retaining the most semantically disruptive outcome at each simulation rollout for efficient local search. Extensive experiments demonstrate that our method achieves state-of-the-art attack performance across both short and long prompts of varying semantics. Furthermore, we find that the fragility of SD models can be attributed to the inherent vulnerability of their CLIP-based text encoders, suggesting a fundamental security risk in current text-to-image pipelines. |
| title | CAHS-Attack: CLIP-Aware Heuristic Search Attack Method for Stable Diffusion |
| topic | Cryptography and Security Artificial Intelligence |
| url | https://arxiv.org/abs/2511.21180 |