"One robot could infect other vulnerable robots nearby ... Attackers could take control of entire fleets of robots."
Summary
A remote code execution vulnerability has been found in Unitree robots, and it is wormable, enabling attackers to infect and control multiple robots.
Similar Articles
Cybersecurity AI: Humanoid Robots as Attack Vectors
This paper presents a systematic security assessment of the Unitree G1 humanoid robot, revealing critical vulnerabilities including BLE provisioning protocol exploits, hardcoded AES keys, and a resident Cybersecurity AI agent capable of exfiltration and offensive operations, arguing for adaptive CAI-powered defenses as humanoids enter critical infrastructure.
Robot Dogs Are a Security Nightmare
This article critiques the deployment of Unitree robot dogs by law enforcement and security firms, highlighting severe cybersecurity vulnerabilities, physical safety risks like LiDAR blind spots, and the reality that many units are remotely operated rather than autonomous.
U of T researchers demonstrate AI worm could target any online device
University of Toronto researchers have demonstrated an AI worm capable of targeting any online device, highlighting a new security vulnerability in AI systems.
Robot ban's cited threat: one person gained control of 7,000 home robots with live cameras
A US determination cites a 2026 Guardian report where one person controlled 7,000 home robots, justifying a ban on foreign-produced advanced robotic devices, while also noting the unreliability of open robot foundation models like LingBot-VLA 2.0.
AI Worms and Viruses Are Coming
Researchers are demonstrating that AI agents can autonomously self-replicate and hack into remote systems, raising concerns about future AI-powered worms and viruses that could evade detection and cause widespread harm.