Feds: AI-Assisted Attacks on Critical PLCs Are Here

When five US federal agencies issued a joint alert on Wednesday, they weren't describing a distant worry or a proof-of-concept experiment. They were describing attacks already underway.

Attackers are using AI-generated exploitation scripts to break into internet-exposed Siemens S7 Series programmable logic controllers (PLCs) at water, manufacturing, energy, and other critical facilities, in what the agencies called an “active threat.” The joint advisory from the National Security Agency (NSA), Cybersecurity and Infrastructure Security Agency (CISA), Federal Bureau of Investigation (FBI), Department of Energy (DOE), and Environmental Protection Agency (EPA) makes explicit that this is not a hypothetical scenario.

AI-Powered Attack Chain

According to the alert, the attackers are combining open source industrial automation libraries – specifically snap7.dll/python-snap7 – with AI coding assistants. This combination lets them craft custom tools that mimic operational technology (OT) monitoring software. These tools provide read/write access to the PLC devices’ memory, configuration data, and ladder logic programs via the S7comm protocol.

“This is not a theoretical risk – it is an active threat,” the feds warned.

The agencies said threat actors are using AI assistance to generate exploitation scripts based on publicly available information about these Siemens S7 Series PLCs. The scripts target initial access, credential access, denial of service, and other objectives. If the PLCs are exposed to the internet or insufficiently segmented, attackers can exploit various critical and high severity known vulnerabilities.

Targeting Critical Sectors

The latest attacks specifically target internet-exposed Siemens S7 Series PLCs across critical manufacturing, energy, water and wastewater, chemical, food and agriculture, and commercial facilities. These are the industries that provide goods and services Americans use every day. The agencies also noted that Siemens S7 Series PLCs are used in the Defense Industrial Base (DIB) and could be targeted there as well.

This appears to be a continuation of the same suite of activity we suspect is affiliated with Iran targeting PLCs, said Cynthia Kaiser, SVP at Halcyon Ransomware Research Center, in comments to The Register. Iran-affiliated actors and adversaries are actively targeting a wide swath of operational technology because these PLCs underpin essential health, safety, and critical infrastructure across society.

Suspected Iranian Connections

While the joint advisory does not attribute the threats to a particular government or criminal group, the agencies’ language points to a broader pattern. The alert follows recent attacks targeting PLCs at water and wastewater facilities across at least 12 states, including a cyberattack that disrupted more than 30 community water systems in Minnesota in late July. Iranian cyber operatives are suspected of being behind those intrusions.

Security experts had previously told The Register that there was no indication the water-system hackers used AI in their intrusions, but they worried attackers would soon add AI to their arsenals. Now, that concern appears to be realized.

AI Lowers the Barrier

Kaiser, a former FBI cyber division deputy assistant director, said the advisory’s focus on AI usage aligns with what experts have expected. State-sponsored adversaries are leveraging AI across the board for discrete tasks, like code checks and scripting, to scale their operations and move faster, she said. The advisory reflects the broader reality that threat actors are using AI to increase their efficiency.

The use of AI indicates an evolution in threat actor capabilities, reducing the need for advanced technical knowledge about OT. It allows attackers to more rapidly develop working industrial control system malware and attack chains.

How Attackers Find Targets

Attackers are using internet-scanning services such as Censys and ZoomEye to find exposed, “poorly protected” PLCs running outdated software or using default passwords. The AI boost makes it easier to turn those scans into working exploits.

Benny Czarny, CEO and founder of critical infrastructure security firm Opswat, told The Register that the bigger issue remains how exposed OT environments are. AI makes it much easier for an attacker to create and modify scripts targeting PLCs, so the barrier to attacking industrial systems continues to fall, he wrote in an email. But the answer is not simply better AI detection.

If data only needs to leave an OT network, use a data diode. There should be no network path back to the PLC for an attacker to exploit.

— Benny Czarny, CEO and founder, Opswat, in an email to The Register

Detection and Mitigation Steps

The feds say critical infrastructure owners and operators should first – immediately – inventory all Siemens S7 Series PLCs in their environment, apply security patches as needed, and make sure no PLCs are accessible from the internet.

They also recommend checking for anomalous S7comm behavior, including connections from non-engineering workstations, unusual data block access patterns, or write operations outside change windows. Sequential IP scanning on port 102 and repeated connection attempts with varying parameters can indicate reconnaissance. Snap7.dll library usage outside approved workstations may also signal intruders on the network.

Owners should use these and the rest of the detection strategies detailed in the government security advisory to hunt for anomalies that may indicate a compromise.

Reducing the Attack Surface

Czarny emphasized that reducing the OT attack surface is critical. If data only needs to leave an OT network, use a data diode, he said. There should be no network path back to the PLC for an attacker to exploit. Yes, AI makes this more urgent. But the real lesson for me is still the same: stop giving attackers a path to the critical system in the first place. And do not rely on antivirus and sandboxes to protect your data flow.

The advisory stresses that these mitigations are not optional. With the threat now active, the window for hardening these systems is closing.

Why It Matters

The combination of AI and exposed industrial controllers could mean more frequent and more sophisticated attacks on the systems that deliver water, power, and manufactured goods. The lowered barrier to entry suggests that even less-skilled attackers could soon attempt similar intrusions, expanding the pool of potential adversaries.

For operators of critical infrastructure, the takeaway is clear: if a PLC does not need to be on the internet, remove it. The longer these systems remain exposed, the greater the chance that AI-assisted attackers will find them first.