Exploiting Vulnerabilities: Universal Adversarial Attacks on Vision-Language-Action Models in Robotics

Abstract:Recently, Vision-Language-Action (VLA) models have revolutionized robotic manipulation by seamlessly integrating visual perception, language understanding, and action generation in an end-to-end learning framework. However, since these models are designed to interact directly with the physical world and humans, their security is critical, and even small vulnerabilities can lead to catastrophic failures. In this work, we propose the Universal Adversarial Object, a sphere with optimized surface texture that significantly degrades task success rates when placed within the robot's field of view. Specifically, our approach introduces a multi-level attack framework that jointly disrupts trajectory planning, task execution, and action control. We validate our method in both simulated and real-world robotic settings. Experimental results demonstrate that the adversarial object reduces the average task success rates by 31.2%-39.9% for two representative VLA models (Pi0 and RDT), with success rates dropping to near zero in complex scenarios. Index Terms--Vision-Language-Action models, adversarial attack, robotic security, universal adversarial object

Accepted to the 2026 IEEE International Conference on Robotics and Automation (ICRA 2026), Vienna, Austria. 8 pages. (c) 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes

Robotics (cs.RO); Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)

arXiv:2609.39178 [cs.RO]

(or arXiv:2609.39178v1 [cs.RO] for this version)

https://doi.org/10.48550/arXiv.2609.39178 arXiv-issued DOI via DataCite (pending registration)