AI Tools Can Speed Up Industrial Exploit Development, Researchers Warn

Security researchers at Forescout have demonstrated that AI tools can be used to quickly adapt existing cyberattack code to work against new targets. In their experiment, they used Anthropic's Claude AI to "port" a known remote code execution exploit from one WAGO programmable logic controller (PLC) model to another, completing the task in hours and for a cost of a few hundred dollars.

PLCs are critical devices used to control machinery in manufacturing, utilities, and other industrial settings. Historically, adapting exploits between different hardware models required specialised expertise and significant time investment, which acted as a natural barrier limiting how many attackers could realistically weaponise such flaws. This research suggests that AI tools can meaningfully reduce that barrier, potentially allowing less skilled attackers to target a wider range of industrial equipment more quickly and cheaply.

While this experiment was conducted by researchers for defensive purposes, it serves as a warning sign for any business relying on industrial control systems, including manufacturers, food processors, utilities, and logistics companies. As AI continues to make offensive cyber capabilities more accessible, organisations using operational technology should treat patching and network segmentation as urgent priorities rather than long-term projects.