# BadBlocks: Low-Cost and Stealthy Backdoor Attacks Tailored for Text-to-Image Diffusion Models

Source: [BadBlocks: Low-Cost and Stealthy Backdoor Attacks Tailored for Text-to-Image Diffusion Models](<https://www.sota2.com/research/paper/68fa7ac31e209131cea850fe>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-30T05:00:11.986Z

Authors: Jia Wu, Yu Pan, Junjun Yang, Yi Du

Published: 2025-08-05T08:48:37.000Z

## Abstract

Despite the remarkable progress of diffusion models in image generation, recent studies reveal their vulnerability to backdoor attacks via covert visual or textual triggers. Although evolving defense mechanisms can detect most existing threats through visual inspection or feature analysis, we introduce BadBlocks-a novel, lightweight, and highly covert attack that challenges these safeguards. By selectively poisoning specific blocks within the UNet architecture while keeping other components intact, BadBlocks requires only 30% of the computational resources and 20% of the GPU time of conventional attacks, effectively democratizing backdoor injection on consumer-grade GPUs. Empirical evaluations demonstrate that BadBlocks achieves a high attack success rate with negligible perceptual quality loss, while successfully bypassing state-of-the-art defenses, particularly attention-based detection frameworks. Layer-level ablation studies further confirm that backdoor mapping does not require full-network fine-tuning, revealing the disparate vulnerability of different neural layers. Overall, BadBlocks significantly lowers the barrier for executing backdoor attacks, presenting a critical security risk. Our code is available at: https://github.com/paoche11/BadBlocks.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Text-to-image Generation on MS-COCO Caption](<https://www.sota2.com/research/sota/text-to-image-generation-on-ms-coco-caption.md>): Text to Image; MS-COCO Caption; ASR: 99.97.

- [Backdoor Attack on Stable Diffusion 20,000 samples](<https://www.sota2.com/research/sota/backdoor-attack-on-stable-diffusion-20-000-samples.md>): Backdoor Attack; Stable Diffusion 20,000 samples; Affected Parameters (M): 18.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2508.03221>)

[PDF](<https://arxiv.org/pdf/2508.03221.pdf>)