AIROSep 3, 2025

ANNIE: Be Careful of Your Robots

arXiv:2509.03383v11 citationsh-index: 8Has Code
Originality Highly original
AI Analysis

This addresses critical safety vulnerabilities in embodied AI systems for robotics applications, representing a foundational rather than incremental contribution.

The paper tackles the security risks of vision-language-action models in embodied AI robots by conducting the first systematic study of adversarial safety attacks, achieving attack success rates exceeding 50% across safety categories and validating results with physical robot experiments.

The integration of vision-language-action (VLA) models into embodied AI (EAI) robots is rapidly advancing their ability to perform complex, long-horizon tasks in humancentric environments. However, EAI systems introduce critical security risks: a compromised VLA model can directly translate adversarial perturbations on sensory input into unsafe physical actions. Traditional safety definitions and methodologies from the machine learning community are no longer sufficient. EAI systems raise new questions, such as what constitutes safety, how to measure it, and how to design effective attack and defense mechanisms in physically grounded, interactive settings. In this work, we present the first systematic study of adversarial safety attacks on embodied AI systems, grounded in ISO standards for human-robot interactions. We (1) formalize a principled taxonomy of safety violations (critical, dangerous, risky) based on physical constraints such as separation distance, velocity, and collision boundaries; (2) introduce ANNIEBench, a benchmark of nine safety-critical scenarios with 2,400 video-action sequences for evaluating embodied safety; and (3) ANNIE-Attack, a task-aware adversarial framework with an attack leader model that decomposes long-horizon goals into frame-level perturbations. Our evaluation across representative EAI models shows attack success rates exceeding 50% across all safety categories. We further demonstrate sparse and adaptive attack strategies and validate the real-world impact through physical robot experiments. These results expose a previously underexplored but highly consequential attack surface in embodied AI systems, highlighting the urgent need for security-driven defenses in the physical AI era. Code is available at https://github.com/RLCLab/Annie.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes