Introduction
My interest in liminal warfare was shaped by David Kilcullen’s articles “The Evolution of Unconventional Warfare” and “Liminal Manoeuvre and Conceptual Envelopment,” as well as his book The Dragons and the Snakes. That interest deepened through observing the growing role of automation and artificial intelligence in the Russo-Ukrainian war, alongside their expanding influence within the United States’ information and security environment.
Through Kilcullen’s work and the rapid development of artificial intelligence (AI), it became clear that modern conflict is no longer defined solely by armies, borders, or kinetic force. Increasingly, it unfolds in the space between recognition and response, between belief and doubt, where perception itself becomes contested terrain. In this environment, artificial intelligence does not merely accelerate warfare—it reshapes how conflict is understood, experienced, and normalized. To grasp what is emerging, we must first distinguish the forms of warfare operating at this threshold.
Liminal Warfare
The primary target of liminal warfare is the thresholds of detection, attribution, and response. Its main domain is the “Gray Zone” between peace and war. The objective is to achieve strategic goals without triggering conflict. Its primary mechanism is to skate around ambiguity, deniability, and incremental actions.
The visibility is deliberately ambiguous or plausibly deniable. Think of a person walking by, minding their own business, but with ill intentions. Key actors are state and non-state actors, proxies, and proxies of proxies working as double agents for a multitude of organizations. When it comes to the tempo, understand that it is gradual, probing, calibrated, and protracted.
The role of artificial intelligence only enhances coordination, attribution denial, and scale. Success is measured by the absence of escalation or by delayed, confused responses that give the actor time to reassess and adapt. Failure collapses ambiguity and risks escalation into open conflict.
Cognitive Warfare
The primary target in cognitive warfare is human perception, cognition, and decision-making. The main domains of cognitive warfare are information, psychology, and perception. The objective is to shape beliefs and behavior to influence outcomes.
The primary mechanisms are narratives, framing, and psychological influence. When it comes to visibility, it is often invisible or normalized within information flows. Key actors are states, non-state actors, platforms, automated systems, etc. The tempo is continuous, adaptive, and rapidly scalable.
The role of artificial intelligence will accelerate narrative creation, targeting, and amplification of the cognitive domain. Success is measured not by fixed metrics, but by shifts in perception, belief, and decision-making. Failure manifests as loss of trust, cognitive fragmentation, and societal polarization.
Bridging the Two
Liminal warfare is the ‘threshold’—the boundary between time and space. When artificial intelligence is applied, the door of perception opens, revealing a kaleidoscope of infinite possibilities. It is not defined solely by overt kinetic violence, but by the ambiguous manipulation of perception, where advantage is exploited and gained before conflict is recognized. Therefore, the focus must be cognitive—for the mind itself is the first line of battle.
Given the immense and nearly limitless possibilities of liminal warfare at both the macro and micro levels, the integration of artificial intelligence allows cognitive warfare to move beyond surface influence and penetrate the cerebral domain—blurring and reengineering the boundaries of reality, reshaping perception to suit the aims of the actor or host, as agency shifts between states, non-state entities, and proxies. So what, then, are its goals?
Instead of targeting military hardware, the objective is to shape perception—creating confusion or division, eroding trust in institutions, and influencing the choices of individuals or entire societies. The “war” is over interpretation and meaning, not territory. But how does artificial intelligence change this?
Why Artificial Intelligence Matters
Artificial intelligence is the game-changer in cognitive warfare because it scales narrative creation and analysis. It can generate text, images, audio, and video quickly and cheaply, producing content that appears highly credible throughout social media. With access to demographics and the vast quantities of behavioral data available online, AI enables messages to be tailored to narrowly defined audiences—by age, location, interests, and disposition. In this sense, AI facilitates liminal cognitive warfare across multiple domains of perception simultaneously.
This capacity enables AI-driven precision targeting. Where human-crafted propaganda was broad and slow, AI can identify cognitive biases, produce compelling content, and automate delivery to those most susceptible to influence. Targeted messaging thus becomes a weaponized precision tool—accelerating narrative dominance while reassuring the audience that nothing is wrong, nothing requires adjustment—the actor controls the transmission. The result is influence that is faster, cheaper, and harder to trace—almost terra incognita cognitiva.
Why this Strategy is Dangerous
“A friend to all is a friend to none,” Aristotle reminds us. The future presents a much grimmer picture: reality for everyone dissolves into no reality at all—spoken now by the ghost in the machine.
For the most part, people can still distinguish what is real. But that margin is narrowing—sometimes slowly, sometimes with startling speed—until the distinction itself becomes difficult to discern. If AI-generated narratives can convincingly mimic authentic content, individuals lose the ability to trust what they see online. The result is not merely erosion, but the undermining of public trust, shared facts, and rational decision-making. Basically, one is left with a form of societal schizoidism—a metaphor for cognitive fragmentation and the loss of a shared reality, a total collapse of trust.
Influence can now be hyper-personalized. AI systems can tailor content based on psychological traits, exploiting specific cognitive vulnerabilities—fear, insecurity, identity—in ways that are difficult for individuals to detect or counter.
There are no borders in AI. Unlike traditional propaganda, it scales instantly and without meaningful constraint. Cognitive warfare is global and continuous, operating 24/7 through social media and messaging platforms; often, all it takes is a nudge. This use-ready capacity does not originate solely from foreign governments—it can be wielded by any actor capable of deploying AI to shape narratives at scale.
Modern media offers a helpful analogy. It increasingly resembles a failed game of telephone. Information moves from source to outlet to outlet, but instead of converging on clarity, it diverges. Those at the event are standing at ground zero, possessing firsthand experience of what occurred. Beyond that zone, information becomes secondary, then tertiary, and distortion begins to accumulate. Each relay introduces new interpretations, biases, and incentives, gradually degrading the message as it spreads.
The key point is that this analogy establishes the problem not as the work of a single bad actor, but as a systemic breakdown in information fidelity. The game of telephone illustrates how cumulative distortion and the loss of original context leave the audience increasingly removed from the source. This creates a quiet storm in which the erosion of trust is structural, not accidental.
What Defending Against Cognitive Warfare Looks Like
Defense is not merely technological; it is intellectual. Narrative intelligence employs tools that detect, analyze, and contextualize narratives in near real time. It focuses on origins, rates of spread, the actors involved, the hosts affected, and the sentiment and impact of the message itself. This AI-assisted analysis reveals who is shaping public discourse—and how.
Transparency and context matter. Exposing the individuals and organizations driving a narrative—who is pushing it, and why—can reduce the effectiveness of manipulative messaging, though it cannot eradicate it. Because the battlefield is the mind, skills such as media literacy, critical reasoning, and fact-checking become defensive assets. Put simply: defense is data + design + education, not censorship alone. Censorship will take care of itself—not as policy, but through social enforcement, as individuals and groups police narratives and impose consequences on those who deviate, pending the next revision of acceptable belief.
The Weaponization of Perception—and How It Is Used
The weaponization of perception and consciousness is nothing new. Throughout history, leaders and their entourages have manipulated information—narratives—to wage conflict not only against external enemies, but against their own populations. Narratives matter because they frame how events are interpreted, determining what is seen, ignored, or believed.
As Mao Zedong once observed, “seal up the enemy’s eyes and ears, and make them blind and deaf… confusing the minds of their commanders and turning them into madmen, using this to achieve our own victory.” The insight here is not merely tactical, but cognitive: the enemy is not only across the battlefield, but within one’s own ranks. This is where narrative power is most decisive.
Narratives shape and regulate a society’s beliefs and behaviors. Artificial intelligence does not invent this dynamic; it amplifies and weaponizes it—making narratives faster, more pervasive, and more ambiguous to counter. Even when a false interpretation is exposed, the critical question remains: how far has it already spread, and how convincing was it to its intended audience?
A widely accepted narrative also serves a secondary function: isolating and marginalizing those who question it. Dissent is not crushed by force, but filtered out cognitively and socially, exposing potential challengers long before they can organize. In this sense, the narrative becomes self-enforcing. Traditional warfare uses tanks; cognitive warfare uses stories.
Strategic Communication
In strategic communication, accuracy is rarely decisive on its own; what matters is how the target audience interprets and internalizes the information. Accuracy informs, but interpretation decides. Even information that is factually flawed or selectively presented can be practical if it anchors itself to a broadly accepted truth, using that credibility as narrative leverage.
The accuracy of strategic silence can be equally deafening. Silence does not simply mean “nothing”; it means “something is missing.” It signals absence, invites inference, and creates an interpretive vacuum that audiences instinctively fill—often with speculation, exaggeration, or worst-case assumptions—rendering even later factual clarification less effective.
In Risk Assessment
When it comes to risk assessment, threat evaluation is no longer limited to kinetic danger; it must also account for the potential for narrative influence. Modern risk assessment increasingly treats narratives as munitions. This shift reflects the reality that physical damage is often secondary to the primary objective: manipulating the population’s perception of reality and its decision-making.
Liminal warfare operates on the “threshold” of detection, using ambiguity to achieve goals without triggering a conventional military response. This ambiguous action allows adversaries to perform covert operations whose sponsorship is suspected but remains unproven, such as Russia’s “little green men” in Crimea.
This pre-maneuver shaping phase—before physical force is employed—is where the battlespace is cognitively conditioned to accept a desired outcome. Success is therefore measured not by territory seized, but by the ability to hijack public attention, normalize ambiguity, and control the narrative.
The best policy to defend against AI-as-a-weapon in cognitive warfare is, obviously, through defense planning. Investment should prioritize narrative intelligence capabilities and training that enable early detection. These capabilities should integrate with existing intelligence, communications, and support structures to identify influence campaigns before they achieve strategic effect. Nevertheless, it still comes down to encouraging critical thinking and verification.
Conclusion
The war for the mind is not new, but artificial intelligence has dramatically altered its scale, speed, and opacity. By accelerating narrative production and exploiting ambiguity, AI intensifies liminal warfare by pushing conflict deeper into the cognitive domain—often before it is recognized as such.
The more disturbing question is not whether cognitive warfare will expand, but how far it can go as agency, interpretation, and meaning are increasingly influenced by artificial systems. In shaping narratives at scale, we are not merely using AI as a tool; we are altering the conditions under which reality itself is perceived and contested. The challenge ahead is both technologically strategic and profoundly human: preserving cognitive autonomy in an environment where perception has become the primary terrain of conflict.
However, a darker question needs to be addressed. How far can cognitive warfare go once artificial intelligence no longer transmits meaning, but inhabits it? Once that point is reached, we are no longer shaping narratives—we are preparing a vessel for a possible influence that does not need to enter the physical world to be real. In other words, Pandora’s box speaks. It is not a prediction. It’s a caution.
Sources
1) Liminal and Conceptual Envelopment: Warfare in the Age of Dragons
Fox, Amos. “Liminal and Conceptual Envelopment: Warfare in the Age of Dragons.” Small Wars Journal, May 26, 2020. https://smallwarsjournal.com/2020/05/26/liminal-and-conceptual-envelopment-warfare-age-dragons/
2) China’s Evolving Military Strategy (Book)
McReynolds, Joe, ed. China’s Evolving Military Strategy. Washington, DC: Jamestown Foundation / Brookings Institution Press, 2017. https://www.google.com/books/edition/China_s_Evolving_Military_Strategy/7WxADwAAQBAJ page 174.
3) Cognitive Warfare: The Fight for Gray Matter in the Digital Gray Zone
Cheatham, Michael J., Angelique M. Geyer, Priscella A. Nohle, and Jonathan E. Vazquez. “Cognitive Warfare: The Fight for Gray Matter in the Digital Gray Zone.” National Defense University Press, 2023. https://ndupress.ndu.edu/Media/News/News-Article-View/Article/3853187/cognitive-warfare-the-fight-for-gray-matter-in-the-digital-gray-zone/
4) Claverie & du Cluzel — The Cognitive Warfare Concept
Claverie, Bernard and François du Cluzel. “The Cognitive Warfare Concept.” Innovation Hub – ACT, 2023. PDF. https://innovationhub-act.org/wp-content/uploads/2023/12/CW-article-Claverie-du-Cluzel-final_0.pdf
5) Kilcullen— Liminal Manoeuvre and Conceptual Envelopment: Russian and Chinese Non-Conventional Responses to Western Military Dominance since 1991 Liminal Manoeuvre and Conceptual Envelopment: Russian and Chinese Non-Conventional Responses to Western Military Dominance since 1991. Issue 2, Online Journal, Queen’s University, 2020. PDF. https://www.queensu.ca/psychology/sites/psycwww/files/uploaded_files/Graduate/OnlineJournal/Issue_2-Kilcullen.pdf
6) Kilcullen — The Evolution of Unconventional Warfare
Kilcullen, David J. “The Evolution of Unconventional Warfare.” Scandinavian Journal of Military Studies 2, no. 1 (2019): 61–71. doi:10.31374/sjms.35. https://www.researchgate.net/publication/333222899_The_Evolution_of_Unconventional_Warfare