Chinese demonstration shows how dangerous commercial robot hacks can be

Researchers are making headway in the "terrifying robot apocalypse" department.

Yicaiglobal reported (via Interesting Engineering) that specialists from the Chinese cybersecurity research group Darknavy recently showed off a method of compromising commercial humanoid robots that has somewhat terrifying implications if used in a real-world setting. At GEEKcon in Shangai, the researchers took a commercially available Unitree robot and demonstrated that it could be hacked with voice commands and, if that wasn't bad enough, used to infect other nearby units.

The robot in question ran off an internal AI agent. By exploiting a flaw in the software, the researchers were able to take over the robot while it was connected to a network, at which point the researchers had the robot use local wireless communication to spread the hack to another nearby robot that actually wasn't even connected to the network at the time. Spreading the hack from one robot to another only took a matter of minutes. Even worse, the researchers were able to issue a command for the robot to physically strike a mannequin on stage.

The implications here are as obvious as they are scary. For decades, cybersecurity risks have largely involved stealing money or information from people. If humanoid robots become commonplace in homes (for use in elder care or things like that), it suddenly opens up a whole new world of horrible possibilities for hackers to commit legitimate physical harm to innocent people. Beyond that, it could disrupt robotic labor operations or critical infrastructure.

For now, let's just be glad all of this was done in a safe environment, with the intent of finding ways to fix these problems rather than propagate them.