- Attackers can remotely execute commands on vulnerable industrial robots without requiring authentication
- Outdated factory robots can expose entire manufacturing networks to devastating cyber attacks today
- Poor network segmentation can enable compromised workstations to hijack nearby collaborative robots
A critical command injection vulnerability has been discovered in Universal Robots PolyScope 5, the operating system that powers the company’s collaborative robots.
The bug, tracked as CVE-2026-8153, has a CVSS score of 9.8 and affects all software versions prior to PolyScope 5.25.1.
An unauthorized attacker who can reach the Dashboard Server network port can make commands that are executed directly on the robot’s underlying operating system.
Command injection vulnerability actually works
This vulnerability could lead to complete compromise of the robot controller, affecting the confidentiality, integrity and availability of the entire system
The dashboard server accepts user-controlled input and sends it to the operating system without properly neutralizing special command elements.
This oversight allows an attacker to inject arbitrary commands that the bot will execute with full system privileges.
The bug was discovered and reported by Vera Mens of Claroty Team82, who coordinated the disclosure through CISA and CERT/CC’s VINCE platform.
Universal Robots has released a patch in PolyScope 5.25.1, which is available on the company’s support page for all affected customers – but the patch does nothing until someone actually installs it, and every day that goes by without an update is another day for attackers to exploit known vulnerabilities.
Therefore, the company strongly recommends that every user update to version 5.25.1 or later as soon as possible.
Network security is the real protection against this exploit
Remote exploitation of this vulnerability requires the bot’s Dashboard Server to be enabled in the user interface and its network port to be accessible to the attacker.
Universal Robots stated that its products are not designed to be accessible directly from the Internet, and direct inbound Internet access is typically prevented by company firewalls.
However, bots accessible from a local network may be vulnerable to attacks originating from that network.
“The security of your network is critical to the security of your robot,” the company warned in its advisory to customers and integrators.
No known public exploits specifically targeting this vulnerability have been reported to CISA at the time of this publication.
This vulnerability is serious and the conditions for exploitation are not difficult to imagine in real industrial environments.
A compromised workstation on the same factory network can easily reach a robot’s Dashboard Server port if the correct network segmentation is missing.
Their behavior afterwards can be unpredictable because it is controlled by people other than its owners.
Therefore, this is unlikely to lead to some kind of autonomous robot revolution, but only represents the preponderance of hackers trying to gain control of the systems.
The rise of collaborative robots that work alongside humans makes this threat particularly worrisome because a compromised robot could cause physical harm to nearby personnel.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds.
