Skip to content

FreezeVLA: Action Freezing Attacks against Vision-Language-Action Models

FreezeVLA: Action-Freezing Attacks against Vision-Language-Action Models

arXiv· arXiv:2509.19870· Xin Wang, Jie Li, Zejia Weng, Yixu Wang, Yifeng Gao, Tianyu Pang, Chao Du, Yan Teng, Yingchun Wang, Zuxuan Wu, Xingjun Ma, Yu-Gang Jiang·Indexed by this site · 23 days ago· Originally publishedAttacks
AI summary

The researchers propose FreezeVLA, an attack framework targeting vision-language-action (VLA) models, in which adversarial images can cause VLA models to "freeze" and ignore subsequent instructions, achieving an average attack success rate of 76.2% across three state-of-the-art VLA models and four robotic benchmarks, significantly outperforming existing methods. The framework generates and evaluates action-freezing attacks through min-max bilevel optimization; the generated adversarial images exhibit strong transferability, and a single image can reliably paralyze the model under multiple language prompts. The authors note that such attacks disconnect the robot's "digital brain" from physical actions and may lead to inaction at critical moments of intervention, highlighting the urgent need for robust defense mechanisms.

Related events and news

1 events · 1 news mentions

News that mentions this paper

VicOne测试:被操纵输入可改变机器人行为

Read originalarxiv.org