2 Colorado Water Utilities Hit by Foreign PLC Tampering

For industrial control system defenders, the Colorado disclosure is a concrete data point in a documented foreign campaign against water-sector OT. The Aug. 19 FBI-NSA-CISA advisory warned of threat actors using AI-generated exploitation scripts against internet-facing Siemens S7 PLCs, and within weeks two utilities reported unauthorized equipment control changes. The event highlights the gap between federal warnings and operational defense at small, resource-constrained utilities.

Foreign actors accessed computer systems at two small private water utilities in Colorado in late August 2026, changed equipment controls, and were subsequently shut out by operators, according to statements from Gov. Jared Polis's office. The disclosure, reported on Sept. 21 and 22, is significant not because of immediate public harm—officials said treatment processes and water quality were unaffected and the utilities served fewer than 200 people—but because it provides another confirmed data point in an escalating campaign against U.S. water and wastewater systems by foreign adversaries. Ally Sullivan, a spokeswoman for the governor, said the Colorado Department of Public Health and Environment followed up with the providers to verify the issues were resolved. The governor's office said it could not confirm which foreign actors were responsible but pointed to CISA's awareness of ongoing efforts by an Iranian-backed group to access drinking water and wastewater systems.

“

Ally Sullivan, a spokeswoman for the governor, said the Colorado Department of Public Health and Environment followed up with the providers to verify the issues were resolved.

The timing aligns with an Aug. 19 advisory from the FBI, NSA, CISA, and other agencies that warned of an active cyber threat to Siemens S7 Series programmable logic controllers used in water systems. Federal authorities described threat actors conducting reconnaissance and capability development against internet-connected Siemens PLCs running outdated software, using AI-generated exploitation scripts disguised as legitimate monitoring tools. That advisory indicated the attackers were specifically probing U.S.-based installations, which makes the Colorado intrusions a worrying validation of the intelligence picture. Although the affected utilities are small, their OT environments likely share the same characteristics the advisory highlighted: limited cybersecurity staffing, legacy and internet-exposed PLCs, and insufficient segmentation between IT and operational networks.

For critical infrastructure defenders, the incident is a textbook example of the cyber-physical convergence problem. Unauthorized actors were able to reach equipment controls—not just data—and alter them before operators manually restored normal operations. This means the intrusion had potential physical consequence, even if no water quality or service impact resulted. The fact that treatment processes were unaffected suggests the attackers either did not target chemical dosing or process integrity, or were interrupted before they could. But the ability to change equipment controls at all indicates that the utilities lacked the network segmentation, access controls, or monitoring needed to prevent or immediately detect operational tampering. U.S. water utilities remain among the least-resourced critical infrastructure sectors, and small private systems like those in Colorado are especially vulnerable because they often operate with minimal dedicated cybersecurity personnel and rely on remote-access tools for maintenance.

The foreign actor attribution remains unconfirmed. The governor's office explicitly declined to identify the actors, and the sources do not independently verify state sponsorship. The reference to an Iranian-backed group reflects CISA's broader national warning rather than a specific determination in this case. However, the pattern is consistent with prior Iranian state-aligned cyber operations against U.S. water and wastewater infrastructure. Those operations have historically focused on programmable logic controllers and human-machine interfaces because many water systems leave these devices directly connected to the internet with default or weak credentials. The use of AI-generated scripts to disguise malicious activity as legitimate monitoring traffic also marks a tactical evolution that complicates signature-based detection and increases the speed at which attackers can scale reconnaissance.