Cognitive Warfare: When Beijing's Weapon Isn't a Missile, It's Your Brain

Cognitive warfare is a broader concept than "information warfare." Information warfare targets messages themselves — circulating false information, suppressing true information. Cognitive warfare goes a step further: its target is the human cognitive process itself, including perception, understanding, judgment, and even willpower.

PLA strategic documents call this "public opinion warfare, psychological warfare, and legal warfare" — the "three warfares," a concept written into the Political Work Regulations as early as 2003. ^[5] The core logic of the three warfares is this: before armed conflict even begins, cause a collapse from within the adversary's society, government, and military — make the adversary's society believe the fight isn't worth fighting, plunge the adversary's government into decision-making paralysis, and make the adversary's soldiers doubt the legitimacy of their own mission.

For Taiwan, this is in fact something happening around the clock, 365 days a year.

Taiwan is currently one of the democracies with the highest density of information-manipulation attacks in the world. According to the annual "Digital Society Project" report published by the Stockholm-based V-Dem Institute, Taiwan has ranked among the top three targets of "foreign information manipulation" for years running. ^[6] More critically, the methods of attack are evolving rapidly: from the relatively crude fake-news websites of earlier years to today's AI-synthesized audio and video, precisely micro-targeted advertising, and "half-true information" that mixes fact with fiction — making the fight against it much harder.

To understand the threat of cognitive warfare, one first has to understand what tools the attacker uses.

Layer one: infiltration of traditional media. The Chinese Communist Party has long influenced the editorial stance of some Taiwanese media outlets through capital infiltration, advertising pressure, and personnel placement. This is not news: since the early 2010s, related investigations and legislative hearings have accumulated substantial evidence that specific television channels' coverage of cross-strait sensitive issues aligns closely with Beijing's official position, though it cannot be traced to any single, explicit chain of control. ^[7] This "ambiguous infiltration" makes the problem harder to address — it is not disinformation, but a real, existing framework of biased reporting.

Layer two: social media influence operations. Platforms such as Facebook, YouTube, and X (formerly Twitter) periodically publish transparency reports disclosing networks of accounts engaged in "coordinated inauthentic behavior" (CIB). Among the account networks that have been taken down, the share of China-origin operations targeting Taiwan ranks among the highest in the world. ^[8] What these operations have in common is that they don't directly spread obvious fake news — instead, they heavily amplify existing controversies and manufacture a sense of social antagonism, making Taiwanese people feel that "we don't even trust each other anymore."

Layer three: the special vulnerability of the LINE ecosystem. Taiwan is one of the markets with the highest density of LINE use in the world, and LINE groups have one key characteristic: they are closed and opaque, making it hard for algorithms to intervene. Once disinformation enters a LINE group tied to a specific age group or geographic community, it can spread rapidly in ways the platform finds hard to detect. According to a Reuters Institute survey, LINE is in fact the channel through which Taiwanese people most often encounter disinformation — as high as 61%. ^[4] Groups with a higher proportion of older members are especially vulnerable: they have had less media-literacy training, place greater trust in content forwarded by people they know, and face higher costs in verifying it.

Layer four: the weaponization of deepfake technology. AI-generated content (AIGC) has sharply lowered the cost of producing cognitive warfare material. Before 2024, making a video of a political figure convincing enough to pass as real required substantial computing power and technical staff; after 2025, open-source video-synthesis tools have lowered the bar to the point that ordinary people can use them. ^[1] What makes it worse is the "fog effect" of deepfakes: even after content is debunked, the doubt it creates can persist. The state of mind of "I'm not even sure whether that clip is real or fake" is already enough for Beijing's purposes.

Layer five: academic and exchange channels. A less-discussed but equally important avenue of infiltration. Through Confucius Institutes, academic exchange programs, and organizations affiliated with the United Front system, Beijing has long cultivated "influential voices" inside Taiwan's universities, research institutions, and religious communities. These voices don't have to be Beijing's direct mouthpieces — as long as, at critical moments, they tend to put forward a narrative of "why confront, what's wrong with talks," the effect of cognitive warfare is already achieved.

A common misunderstanding is that the goal of cognitive warfare is to "make Taiwanese people support unification" or "make Taiwanese people believe in the 1992 Consensus." This understanding is wrong, or at least too shallow.

PLA strategic research shows that in the pre-war phase of cognitive warfare, the goals are far more specific:

Lowering the will to defend. Make Taiwanese people feel that "resistance is pointless," "the United States won't come," "if it comes to fighting, only Taiwanese people will die." There is no need to convince anyone that "unification is good" — it is enough to make enough people question whether defense is even worth it. ^[5]

Paralyzing the decision-making center. In a crisis, plunge the Taiwanese government into a fog of "is this disinformation or is it real," delaying mobilization and response time. Even a delay of just six to twelve hours could be militarily disastrous.

Fracturing social consensus. Taiwanese society already has political divisions along blue-green, generational, and urban-rural lines. The work of cognitive warfare is to amplify and inflame these divisions, so that when facing an external threat, Taiwanese people are still busy tearing each other apart internally instead of coming together.

Isolating Taiwan's international narrative. Through opinion-shaping efforts that influence media and public opinion in third countries, give the narrative that "a Taiwan Strait conflict would be Taiwan's own provocation" a degree of circulation in Southeast Asia, Latin America, Africa, and elsewhere, so that potential supporters of Taiwan face domestic opposition to that support.

What makes this set of goals so formidable is its "low bar for success": it doesn't need everyone to believe it — it only needs to confuse enough people, and cognitive warfare has already achieved its tactical effect.

Facing this kind of attack, Taiwan is not defenseless. Taiwan is, in fact, one of the democracies in the world with the most real-world experience resisting cognitive warfare.

A civil-society fact-checking ecosystem. Organizations such as the Taiwan FactCheck Center (TFC), MyGoPen, and CoFacts form a multi-layered civic fact-checking network that can, within hours to days of an election or a sensitive event, complete fact-checks of widely spread disinformation and publish the results. ^[3] This ecosystem was built partly with government funding, but relies more on the spontaneous mobilization of civil society, embodying Taiwan's distinctive democratic resilience.

The government's institutional response. The Ministry of Digital Affairs (moda), the Executive Yuan's cyber security office, and the National Security Bureau each have their own responsibilities in defending against cognitive warfare, and in recent years have built intelligence-sharing mechanisms with the EU, Japan, and the United States through the "Democracy Network" program. ^[2] The problem is that coordination mechanisms are still incomplete — real-time cross-agency response still looks slow under real pressure.

Media literacy education. Since 2019, Taiwan has added media-literacy content to its elementary and middle school curricula, making it one of the earliest countries in the world to do so. But the effects of education take time to accumulate, and its penetration among older generations remains far from sufficient.

Even with these lines of defense, Taiwan remains vulnerable on several structural weak points:

Structural weakness one: LINE's regulatory gray zone. The spread of disinformation in closed groups cannot be effectively intervened in by platform algorithms, and it also falls outside the regulatory framework of the current Fundamental Communications Act. How to build an effective, public-health-style intervention mechanism while still protecting communication privacy still has no complete answer.

Structural weakness two: a blind spot in cognitive security inside companies. Most Taiwanese companies' information-security rules focus on cybersecurity (hacking, data leaks), but almost none have a corresponding "cognitive security" mechanism — scenarios such as employees forwarding disinformation on company chat tools, or senior decision-makers being misled by deepfake videos, lack any systematic response procedure.

Structural weakness three: the intersection of military mobilization and cognitive warfare. In a crisis, the most dangerous moment for cognitive warfare is "the first twenty-four hours after a mobilization order is issued." If a flood of half-plausible messages pours in during that window — "the front line has fallen," "the president has fled," "the U.S. fleet is withdrawing" — mobilization efficiency would drop sharply. Based on public information, the degree to which Taiwan currently has contingency plans ready for this intersection scenario still appears insufficient.

Cognitive warfare is not only a problem for the government and the military. For businesses, it is a steadily escalating operational risk.

Decision-making security for senior executives. Deepfake technology can forge the voice and image of partners, clients, and even executives. "CEO fraud" — an evolved version of business email compromise — already has an AI voice-synthesis variant that orders finance staff to make urgent wire transfers. Any decision involving a major movement of funds must have an out-of-band verification procedure.

Supply chain narrative risk. During cross-strait sensitive periods, the cognitive warfare businesses may face includes fake news spread by competitors ("such-and-such company has been sanctioned in Taiwan," "such-and-such brand is complying with Chinese export controls"). This kind of message, circulating in industry group chats, can inflict damage on a company's reputation and client relationships within hours that is hard to reverse quickly. Building a rapid-clarification mechanism is necessary.

Employee information hygiene. Company employees are both targets and potential vectors of transmission. Regular information-hygiene training — not just cybersecurity awareness, but the ability to "identify manipulative content" — should become a standard training course for medium and large enterprises.

The fundamental dilemma Taiwan faces in confronting cognitive warfare is a structural contradiction within democratic institutions themselves: effective countermeasures can, in themselves, damage democratic values.

The government cannot build a censorship mechanism — that is exactly the path Beijing wants Taiwan to go down: a self-limiting democracy. But imposing no restrictions at all, letting information-manipulation operations circulate freely on platforms, likewise damages the quality of democracy.

The path Taiwan is currently taking is, broadly, "accelerated transparency": speeding up fact-checking disclosure of disinformation rather than taking content down; investing in media-literacy education rather than restricting what people can receive; focusing on exposing foreign "coordinated inauthentic behavior" rather than restricting speech itself.

This path is not without cost. It works more slowly than authoritarian-style control, and in the short window when a crisis erupts, it may not be able to keep up with the speed at which disinformation spreads at all. But it preserves the thing that matters most about Taiwan: an open society capable of doubting and correcting itself.

And that is precisely what Beijing least wants Taiwan to have.