cs.RONov 8, 2026

When Coding Agents Get Arms: Safety Risks Across the Digital-Physical Boundary

Authors: Yiming Sun, Mi Zhang

Organizations: Fudan University, JADE AI(Whitzard AI) Team

Abstract

Advances in spatial understanding have enabled frontier models to power coding agents that control a real robot arm zero-shot and end-to-end, without task-specific demonstrations or retraining. This brings two new safety challenges. First, the extensive knowledge of frontier models lets these agents perform tasks that earlier VLA and WAM paradigms could not, opening the door to new forms of capability misuse. Second, coding agents can access digital resources such as web pages and emails. Threats in these resources, such as indirect prompt injection, can carry over into the physical world when the agent controls a robot. We therefore introduce ArmedAgent, the first systematic benchmark to evaluate both risks for coding agents performing zero-shot robot control. It covers robustness, malicious use, and indirect injection across 36 tasks in 5 scenarios, evaluated in simulation and validated on a physical robot arm. A distinctive portion of the tasks target harmful behaviors that earlier VLA paradigms could not perform. Testing several frontier models, including GPT-6-Astra, under widely used agent frameworks, we find substantial safety risks in current systems.

Figures & tables

Explore similar work

CardsList
  1. V3^3AST-Bench: Verifiable, Valid, and Versatile Safety Evaluation for Coding Agents

    Oct 26, 2026Weitao Feng, Hengheng Xiong, Zihao Zeng +6Runtime Verification for AI AgentsAI Agent Safety

  2. LEASH: Stay Safe in Agentic Robotics

    Nov 10, 2026Qinfeng Zhu, Weiguang Zhao, Ke Sun +5Robot SafetyAgentic AI for Robotics