What is Cognitive Warfare?

By Matthew Parish

Friday 2 October 2026

For most of human history, armies have fought over territory. In the twenty-first century they increasingly fight over something rather harder to locate on a map: the contents of the human mind. Ukraine has become one of the principal laboratories in which this new form of warfare is being developed, tested and resisted.

The expression “cognitive warfare” sounds faintly sinister. It evokes laboratories filled with psychologists, intelligence officers and computers manipulating populations through methods invisible to their victims. Like many fashionable expressions in contemporary security studies, it is also sometimes used so broadly that it threatens to mean almost everything and therefore nothing at all.

Nevertheless there is a serious idea behind it. Cognitive warfare describes attempts to influence the way individuals or populations perceive, interpret and make decisions about the world around them. Its objective is not merely to persuade people of a particular proposition. It may instead seek to alter the intellectual and emotional environment within which people decide what they believe in the first place.

This distinction separates cognitive warfare from the simpler forms of propaganda familiar from the twentieth century. Traditional propaganda says: believe this. Cognitive warfare may instead say: believe nothing, distrust everybody, become frightened, become exhausted, become angry with your neighbours, conclude that the institutions surrounding you are incompetent and eventually cease trying to distinguish truth from falsehood.

For Ukraine, this distinction is anything but academic.

The battlefield inside the head

Russia’s full-scale invasion of Ukraine has produced the largest conventional European war since 1945, yet almost every physical battle in this conflict has been accompanied by an informational one. A missile strikes an apartment building and within minutes competing explanations circulate about what happened. A town changes hands and arguments immediately begin about whether the event represents a decisive breakthrough, a tactical adjustment, a catastrophe or something that never happened at all.

The physical event remains the same. What changes is its meaning.

This has always been part of warfare. Julius Caesar wrote accounts of his campaigns that conveniently enhanced his political reputation in Rome. Napoleon understood the value of newspapers, while the governments of the First and Second World Wars maintained enormous propaganda operations. Radio subsequently allowed governments to speak directly across enemy frontiers, and the Cold War produced increasingly sophisticated contests between competing systems of political communication.

What is different today is principally scale, speed and fragmentation. Propaganda was once expensive. Governments needed printing presses, radio transmitters, television stations or newspapers. The internet reduced those costs dramatically, social media reduced them further and artificial intelligence is now pushing the marginal cost of producing persuasive false content towards zero.

A single person can create thousands of accounts. A convincing photograph can be manufactured in seconds. Voices can be cloned, video can be synthesised and armies of automated accounts can repeat the same claim until its sheer ubiquity gives it the appearance of authenticity. The modern propagandist no longer needs to persuade an editor to print a story because everyone has become his own editor and publisher.

This transforms the nature of the problem. The question is no longer merely how governments should rebut false information. It is how societies can continue making rational collective decisions when the information environment itself has become contested terrain.

Russia’s distinctive approach

The Russian state did not invent propaganda, psychological operations or political deception. It has nevertheless developed a particularly sophisticated modern combination of these methods, drawing upon Soviet traditions while exploiting the peculiarities of Western social media and open societies.

One frequently misunderstood feature of Russian information operations is that they do not necessarily seek to make foreigners admire Russia. That would often be an implausible objective. Nor do they always attempt to convince audiences of one coherent Russian account of events.

Quite the reverse may be true. The same event can be surrounded by numerous mutually contradictory explanations. The immediate purpose is not necessarily to establish which explanation is correct but to make certainty itself appear impossible.

This strategy exploits an important weakness in human reasoning. Most people do not have the time, expertise or inclination to investigate every claim they encounter. When presented with one plausible account and one implausible account, they may choose the plausible one. When presented with twenty competing accounts, each accompanied by apparently authoritative photographs, quotations and experts, they may eventually decide that nobody knows anything.

That conclusion is itself politically consequential. A citizen who believes that every politician lies, every journalist is corrupt, every photograph may be fabricated and every institution is secretly controlled by somebody else becomes extremely difficult to mobilise for any collective purpose.

Cynicism can therefore be a weapon.

Ukraine’s particular vulnerability

Ukraine presents an unusually complex environment for cognitive warfare because the country has been subjected simultaneously to physical invasion and continuous information attack. The two interact constantly, and the consequences of a successful information operation may have direct military significance.

Consider mobilisation. Russia does not need to persuade every Ukrainian that mobilisation is illegitimate. She merely needs to magnify genuine anxieties, circulate the most disturbing individual incidents, invent additional incidents where convenient and encourage the impression that everybody except the reader is avoiding military service.

The underlying facts may include genuine problems. That is important because the most effective information operations are rarely constructed entirely from lies. They begin with something true, strip away inconvenient context, add emotionally provocative material and amplify the result until an exceptional event appears universal.

The same technique can be applied to corruption, relations between military commanders, Ukrainian language policy, Western assistance, refugees, relations between Kyiv and local government, battlefield losses and virtually every other subject capable of generating strong emotions. The purpose is not necessarily to invent social divisions but to identify divisions that already exist and deepen them.

This is one reason why simply calling everything “Russian disinformation” is inadequate. A democratic society must retain the ability to distinguish between hostile manipulation and legitimate criticism. Ukrainians are perfectly entitled to criticise their government, military administration, mobilisation procedures and political leaders, and the existence of Russian attempts to exploit such arguments does not make the arguments themselves illegitimate.

Indeed the failure to preserve this distinction can become a vulnerability of its own. If citizens conclude that uncomfortable facts are routinely dismissed as enemy propaganda, trust in institutions declines and the propagandist benefits.

From disinformation to cognition

This brings us to the distinction between countering disinformation and cognitive security.

Disinformation is comparatively easy to conceptualise. Somebody publishes a false claim. Investigators establish that it is false. The government, journalists or independent researchers explain the truth and, in the ideal case, the falsehood disappears.

Reality is rarely so cooperative.

Suppose a fabricated story reaches five million people during its first six hours online. Twenty-four hours later a government agency publishes a detailed and entirely accurate rebuttal that reaches 100,000 people. Technically the disinformation has been disproved. Strategically the operation may nevertheless have succeeded.

Cognitive warfare therefore concentrates upon the entire process by which information enters and influences human decision-making. Who is the intended audience? Why is this particular claim emotionally attractive to that audience? Through which accounts does it travel? Which genuine grievances make it plausible? What behaviour is the operation intended to produce?

Those questions lead far beyond traditional fact-checking. They involve psychology, sociology, intelligence analysis, computer science, advertising, political communication and increasingly artificial intelligence.

The resulting discipline is uncomfortable because it approaches the boundaries between defending a society from manipulation and manipulating the society in order to defend it. Democratic governments therefore face a dilemma that authoritarian governments can largely ignore.

The democratic paradox

An authoritarian state can conduct cognitive warfare with relatively few philosophical inhibitions. It can censor inconvenient information, manufacture favourable narratives, intimidate journalists, operate networks of false identities and conceal the entire apparatus behind national-security secrecy.

A democracy cannot safely imitate these methods without damaging the thing it claims to be defending.

This creates what might be called the democratic paradox of cognitive warfare. The more sophisticated democratic governments become at influencing their own information environments, the more important it becomes that citizens can distinguish legitimate national-security activity from political manipulation.

Ukraine confronts this dilemma under exceptionally difficult circumstances because it is simultaneously fighting for national survival and attempting to remain a pluralistic democracy. Wartime censorship of military information is plainly necessary. Publishing the coordinates of Ukrainian air-defence systems would not constitute admirable journalistic openness.

Yet there is no equally obvious line separating operational security from political embarrassment. Nor is there a simple definition separating Russian amplification of a genuine Ukrainian controversy from the controversy itself.

Institutions responsible for cognitive security therefore require credibility more than almost any other branch of government. Once the public begins to suspect that “disinformation” simply means information inconvenient to the authorities, the institution has compromised its principal asset.

Trust cannot be manufactured by decree.

Artificial intelligence changes everything

Artificial intelligence introduces another dimension to this struggle because it destroys one of the assumptions upon which modern information societies were built: that convincing evidence is expensive to fabricate.

For most of the photographic era, seeing was approximately believing. Photographs could be manipulated and films could be staged, but sophisticated fabrication required expertise, equipment and time. These practical constraints provided a modest barrier against industrial-scale falsification.

That barrier is disappearing.

A photograph showing a politician in a compromising situation can now be generated without the politician ever having been there. A voice recording can reproduce the speech patterns of somebody who never uttered the recorded words. Video synthesis is improving rapidly, while large language models can produce enormous quantities of persuasive prose adapted to different political audiences.

The obvious danger is that people believe fabricated material. The more profound danger may eventually be the opposite: people cease believing authentic material because they know that everything can be fabricated.

This phenomenon is sometimes described as the liar’s dividend. Once convincing fabrication becomes commonplace, somebody confronted with genuine evidence can simply claim that it was generated by artificial intelligence.

The implications for wartime Ukraine are profound. Evidence of Russian atrocities, battlefield events, intercepted communications and political statements increasingly exists within an environment in which technological plausibility alone cannot guarantee authenticity.

Verification therefore becomes more important at precisely the moment when verification becomes more difficult.

The algorithms know us

There is another reason why cognitive warfare is different from traditional propaganda. Contemporary information systems do not merely distribute propaganda more efficiently; they reveal which propaganda works.

Social-media platforms continuously collect behavioural information. They know which stories users open, how long they watch a video, what makes them angry, which accounts they follow and which subjects cause them to comment or share material.

This creates the possibility of propaganda optimised through experimentation.

A hostile information operation can publish numerous versions of essentially the same narrative and observe which receives the greatest engagement. Successful variants can then be amplified while unsuccessful ones disappear. The process begins to resemble evolutionary selection, with narratives competing for survival inside an algorithmic ecosystem.

Artificial intelligence can accelerate this process dramatically. Instead of producing one propaganda message for an entire country, an information operation can theoretically generate different versions for different demographic, linguistic, geographic or political audiences.

The propaganda poster addressed everybody who walked past it. The cognitive operation may eventually address you.

Can cognitive warfare actually control people?

At this point some scepticism is necessary. Security terminology has a tendency towards exaggeration, and “cognitive warfare” can easily become a fashionable expression implying powers that governments do not actually possess.

Human beings are not programmable machines. Exposure to propaganda does not automatically determine behaviour, and decades of research into political persuasion demonstrate that people’s beliefs are shaped by families, communities, identities, experiences and pre-existing convictions as well as by media consumption.

Nor should every surprising political development be attributed to mysterious foreign manipulation. People can become angry with governments without Russian assistance. Elections produce unexpected results because voters possess agency, while social divisions frequently arise from genuine conflicts of interest.

The concept of cognitive warfare becomes dangerous when it transforms ordinary disagreement into evidence of enemy activity. A government that begins viewing every critic as a potential vector of hostile influence has ceased defending democratic society and started undermining it.

The correct objective is therefore resilience rather than control.

Building a society that is difficult to manipulate

A cognitively resilient society is not one in which everybody believes the government. That would be authoritarianism, not resilience. It is one in which citizens possess enough confidence in reliable institutions and enough critical judgement to navigate uncertainty without becoming paralysed by it.

Independent journalism is part of that defence. So are transparent government communications, functioning courts, credible public institutions and education that teaches people how to evaluate sources rather than merely what conclusions they should reach.

Government agencies have a role as well. They can identify coordinated foreign influence operations, expose fabricated material rapidly, cooperate with foreign governments and technology companies and explain the methods through which hostile networks operate.

Their effectiveness nevertheless depends upon restraint. Every time an authority describes a genuine controversy as disinformation merely because the controversy is inconvenient, it spends some of its credibility. In cognitive warfare, credibility is ammunition that cannot easily be replenished.

Ukraine has learned this under the harshest conceivable conditions. Its population has demonstrated remarkable resilience against years of Russian predictions that the state is about to collapse, the army is about to disintegrate, Western assistance is about to disappear or the country’s political leadership is about to flee.

Yet resilience is not permanent. People become tired, wars continue longer than expected and repeated exposure to uncertainty corrodes confidence. The information battlefield therefore becomes more important rather than less important as a prolonged war continues.

The battle after the battle

This helps explain why the language of “cognitive confrontation” now appears in discussions about Ukraine’s national-security institutions. The problem is no longer simply identifying fake news. It is understanding how hostile states attempt to alter the perceptions through which societies interpret reality.

The implications extend far beyond Ukraine. Every European democracy now operates inside substantially the same technological information environment, even if most enjoy the luxury of doing so without Russian missiles falling upon their cities. Elections, public-health emergencies, migration disputes, economic crises and wars can all become targets for external information operations.

Ukraine may therefore be several years ahead of its European partners in confronting a problem that will increasingly become theirs as well. That does not mean every Ukrainian experiment should be copied, nor that the vocabulary of wartime security should casually migrate into peacetime democratic politics.

It does mean that the Ukrainian experience deserves close study.

The central lesson may ultimately be surprisingly old-fashioned. Cognitive warfare is fought with artificial intelligence, algorithms, social networks, behavioural data and technologies that would have appeared magical a generation ago, but its decisive resource remains trust between human beings.

Russia can manufacture photographs, accounts and narratives. Artificial intelligence can multiply them almost without limit. No technology, however, can manufacture genuine institutional credibility quite so easily.

The most successful defence against cognitive warfare may therefore consist neither of censorship nor of an ever-larger bureaucracy devoted to correcting the internet. It may consist of building public institutions that tell the truth often enough that people continue listening when those institutions say something important, alongside independent media sufficiently rigorous that their readers know the difference between verified information and rumour.

Ukraine is fighting for territory, sovereignty and survival. At the same time she is fighting a quieter war over whether Ukrainians, Europeans and the wider world can continue agreeing upon enough facts to understand what that first war is actually about.

The weapons used in this second conflict may be invisible, but that does not make their consequences imaginary. In an age in which almost any image can be fabricated, almost any voice can be imitated and almost any story can be multiplied a million times before breakfast, preserving the capacity to distinguish reality from manipulation is becoming part of national defence itself.