Robots Can Be Controlled by Spoken Commands

Cybersecurity specialists from DARKNAVY revealed vulnerabilities in humanoid robots that can be exploited via spoken commands. In a test, they successfully hijacked a commercially available robot, which then transmitted the exploit to another robot, demonstrating a cascading breach. The researchers showcased the potential danger by directing the hijacked robot to strike a mannequin, highlighting the risks of physical harm. This incident underscores the need for improved security in AI-driven systems.

Want More Context? 🔎