Cognitive warfare in OT security: The hidden risk

Every major OT security investment your organization has made is built on a shared assumption: the adversary wants your systems. You’ve invested in firewalls, network segmentation, endpoint detection, and internal monitoring. Controls designed to prevent unauthorized access and stop damage once someone gets in.

That assumption is no longer sufficient. Nation‑state actors have confirmed access inside U.S. critical infrastructure today. Known targets include energy, water, transportation, and defense industrial base environments. In many cases, these actors remain inside networks for years without disrupting operations or exfiltrating data. No alarms. No outages. Just presence.

What they learn during that time is not only technical. They learn who your operators are, who they trust, which alarms are routinely ignored, how information moves during shift change, which vendor advisories are acted on without scrutiny, and which senior operator everyone calls when something feels wrong.

This is not reconnaissance for a technical attack. It is reconnaissance for something most security programs do not account for at all.

Cognitive Warfare: An overlooked risk in OT security

Cognitive warfare targeting industrial operators is the deliberate manipulation of human decision‑making to produce physical outcomes. An adversary does not need to compromise your systems if they can compromise the people running them.

What this looks like in practice is rarely obvious and almost never technical: 

• An operator subtly conditioned over months to distrust a specific safety system. 

• A shift handover where false or incomplete information entered the record weeks earlier and was never corrected. 

• A vendor advisory that influenced a configuration decision in the exact direction the adversary needed. 

Note: None of these actions trigger a SIEM alert. None appear in a penetration test. 

Existing OT security frameworks like NERC CIP, IEC 62443, and NIST 800‑82 don’t meaningfully address this threat. They treat operators primarily as sources of unintentional error to be minimized, not as deliberate targets of adversarial activity. As a result, they fail to account for the manipulation of operator trust, judgment under pressure, long‑term shaping of decision‑making behavior, and abuse of legitimate operational processes and information flows. Operator judgment is an adversarial attack surface. This is the gap.

Now add AI to the equation

Until recently, cognitive operations were constrained by time and resources. Mapping trust relationships, learning a facility’s culture, and crafting technically credible influence required months of effort by skilled analysts.

AI removes those constraints. Pattern recognition across communications data can surface operator trust networks in hours. Language models can generate content that mirrors a facility’s technical language, documentation style, and operational context with a level of consistency and scale humans cannot sustain. What once required a dedicated team now requires a capable model and sufficient data.

This threat is not theoretical. It is present, and it is becoming faster and cheaper to execute. 

What to do about it

Defending against this class of threat requires capability across five areas: 

• Recognition: Understanding what cognitive and influence campaigns look like in OT environments. 

• Information environment hardening: Protecting the integrity of the information operators rely on to make decisions. 

• Detection: Identifying campaigns through behavioral and decision‑making signals rather than purely technical alerts. 

• Governance: Controlling how information enters operational workflows and which channels it is allowed to travel through. 

• Response: Defining a process for investigation, containment, and recovery when a campaign is suspected or confirmed. 

None of this replaces your existing technical security investments. It complements them. A capable adversary that cannot defeat your technical controls will look for a way around them. Your operators are that path. 

Takeaway

Organizations that build this capability early gain a compounding advantage. Behavioral baselines, trusted reporting, and meaningful detection all take time to mature through real operational use. This is not something that can be stood up in response to a threat advisory or rushed into place during an incident. 

Recent government advisories, including those on Volt Typhoon, confirm why this matters. Nation‑state actors are already maintaining long‑term, silent access inside critical infrastructure, not to disrupt, but to wait, observe, and position themselves. 

The question is no longer whether this threat exists, but whether organizations are prepared to defend more than just systems. Those that fail to protect the people and decisions that keep operations running will never see the attack coming. 

---

The shift cannot wait. Start by evaluating how your OT environment manages operational trust, decision-making, and information flow through our OT Security Assessment grounded in real-world operations. Learn more.