A string of security weaknesses in areas such as default configurations, authentication mechanisms and open source components could enable cybercriminals to easily take over robots used in industrial settings, researchers warned.

An analysis of major industrial and collaborative robots, or cobots, by IOActive revealed close to 50 vulnerabilities that, if exploited, could harm the people who work with them. The firm created a series of videos to demonstrate what tampering with cobots could look like, including swinging robotic arms that have had safety features and emergency settings disabled.

Industrial Cobots Put Workers at Risk

The general public might not be familiar with cobots, but they are far more advanced than you might expect. Companies such as Rethink Robotics, Baxter/Sawyer and Universal Robots have designed cobots to assist human employees with various tasks, using microphones and cameras to see and hear, SecurityWeek reported. That potentially makes them even more dangerous if threat actors hijack them for malicious purposes.

Some of the biggest security holes the researchers discovered involve the way industrial cobots communicate, according to The Daily Mail, as well as poor password protection and lack of authentication. The research concluded that, in some cases, cybercriminals could gain remote access privileges with relative ease, giving them the ability to crush human skulls with a mechanical arm, for example.

Protecting Cobots From Cybercriminals

Bloomberg contacted several of the cobot manufacturers called out in the research, some of which said they were aware of the findings and were already working on fixes. A few did not respond or were unable to confirm whether all the potential holes had been patched. Given how widely cobots are being deployed in many industrial environments, however, there may be more pressure from customers to ensure that they’re safe from cybercriminals.

Even if cobots aren’t manipulated to attack human beings, they could also be used to spy on organizations, Threatpost suggested. Of course, there have been no reports of any such incidents yet.

As more companies are starting to pay attention to security risks associated with the Internet of Things (IoT), cobots may become another endpoint that requires increased vigilance as threat actors learn more about the potential to turn them to their advantage.