China has been increasing pressure on Taiwan below the threshold of a traditional military attack throughout the past decade. These gray zone tactics include diplomatic, economic, and security measures as well as the deployment of ships and aircraft across the boundaries delineating Taiwan’s maritime zone and airspace. While each action is small in scope, over time, they build pressure against Taiwan and are shaking the democratic foundation of the island without China firing a single shot. Central to this campaign is China’s strategy of cognitive warfare. 

Understanding Cognitive Warfare

In the 1990s and early 2000s, China underwent a revolution in military affairs that modernized the People’s Liberation Army. During this phase of military progress, China developed strategies to wage what it termed unrestricted warfare (超限战), utilizing domains outside of conventional warfighting to gain strategic advantages. Alongside China’s development of unrestricted warfare, scholars in the West also formulated new notions of power, notably hard and soft power. Hard power is conventional military might that uses force to coerce an enemy. Soft power consists of using tools such as media and public diplomacy to attract and persuade the target’s population. Then there is sharp power, which can appear similar to soft power by way of mechanism, but with an intent to confuse and demoralize rather than persuade. Within both unrestricted warfare and sharp power lies information warfare—when a government deliberately uses information networks (that is, the Internet) to gain a competitive advantage. This can take many forms, from traditional propaganda campaigns to social media bots. Since information warfare is below the threshold of a military assault, yet is not behavior typically seen in peacetime, it is also considered a gray zone tactic. The most modern and possibly most disruptive form of information warfare is cognitive warfare.

The term cognitive warfare came into play in 2017 to describe the tactic of manipulating an enemy’s cognition. What distinguishes cognitive warfare from other information-related strategies is that the goal is not necessarily to make an enemy believe certain information. While cognitive warfare does utilize disinformation, the intended effects are instead to impede decision-making, polarize society, weaponize identity and narratives, and erode the will to fight. For example, Chinese-controlled bots on Taiwanese social media are less concerned with selling single, pro-China narratives than with sowing distrust and driving a wedge between the Taiwanese people and their government. By eroding societal trust and faith in government, China can degrade Taiwan’s will to fight and encourage reunification. Chinese cognitive warfare tactics flare up during election cycles, specifically targeting the independence-leaning Democratic Progressive Party (DPP) with rumors and conspiracies. 

Beijing’s Efforts to Influence Taiwan’s Election Process

According to Doublethink Lab, a Taiwanese civil society organization that studies disinformation, China sought to influence Taiwan’s 2020 election results through propaganda and content farms. The lab’s 2021 report, “Deafening Whispers,” details various actors involved in China’s information manipulation and their roles. China’s Propaganda Department, People’s Liberation Army, and United Front Work Department (UFWD), among other entities directly controlled by the Chinese Communist Party (CCP), set the tone and messaging of the narratives that get pushed out. However, they are rarely the ones spreading the propaganda, with the Chinese People’s Political Consultative Conference and the Taiwan Affairs Office directing content farms to disseminate disinformation. Doublethink Lab’s research highlights the decentralized nature of Chinese disinformation campaigns. Over the past six years, the actors involved have stayed the same, while the technology to spread disinformation has evolved.

Doublethink Lab’s 2023 and 2024 reports, which analyzed the 2022 local and 2024 presidential elections, noted several changes from previous election cycles. Starting with the 2022 local elections, China relied on fewer content farms as social media platforms increased efforts to stop coordinated inauthentic activity. With the proliferation of artificial intelligence (AI) leading up to the 2024 presidential elections, disinformation transitioned from text-based posts to AI-generated or -manipulated videos. Instead of using content farms to produce low-quality disinformation articles, Chinese bots increasingly focus on amplifying already existing stories. For example, in December 2021, the rumor began circulating on social media that the United States is hollowing out Taiwan’s leading semiconductor producer, Taiwan Semiconductor Manufacturing Company, and will abandon Taiwan’s defense once semiconductor production moves to the United States. While this narrative did originate in Taiwan, UFWD-linked media groups quickly posted exaggerated accounts claiming that the ruling DPP is selling off semiconductor manufacturing to satisfy its own interests. Chinese entities were not attempting to invent a false story, but rather took advantage of Taiwan’s open society to exaggerate already existing fears. Current Chinese disinformation campaigns focus less on selling particular stories and more on stoking partisan divisions and eroding trust in government. While it is difficult to prove that China’s campaigns have caused a change in sentiment, Doublethink Lab has proved a correlation. 

A Doublethink Lab phone survey measured Taiwanese citizens’ political affiliations and beliefs and compared them to how they receive their information. China’s information manipulation targets people aligned with either the opposition Kuomintang (KMT) or the Taiwan People’s Party (TPP) to spread anti-DPP messaging. While partisan disagreements online are a staple in any functioning democracy, Chinese bots can often be distinguished by synchronized post timing, view-botting, and phrases distinct to mainland China. Chinese-linked accounts on Facebook and X infiltrate online KMT and TPP echo chambers to spread disinformation regarding Taiwan, DPP leadership, and cross-strait relations. These narratives first get spread within echo chambers before reaching a broader audience. When surveyed, both KMT and TPP supporters tended to agree more with Chinese-linked rumors, while DPP supporters did not. Additionally, those who stated they received information from social media influencers, online content creators, and fan pages, all of which have been highly targeted by Chinese campaigns, tended to agree more with China’s narratives. While causation cannot be established, the strength of the correlation is significant. What makes China’s current method of information manipulation unique, and what characterizes it as cognitive warfare, is that the end goal is not necessarily to make the audience believe a certain story. Instead, China focuses its efforts on distorting the Taiwanese public’s opinion of the DPP and eroding trust in their government. Each narrative or conspiracy is not significant in itself, but their cumulative effect undermines the legitimacy of political debates in Taiwan. 

While spreading disinformation on social media is a central part of China’s strategy, there are other significant means of information manipulation. At a congressional hearing, DPP legislator Puma Shen testified that China uses media, academia, technology, diplomacy, and the military to interfere in foreign countries’ information environments. Shen describes three strategies of spreading disinformation. The first, direct information manipulation, includes the CCP’s propaganda department and social media campaigns. The second category is indirect investment, when China provides financial backing to marketing companies and influencers that spread disinformation. Taiwanese media companies that partner with Chinese corporations for financial support also often self-censor on issues critical of the CCP. The third approach is ideology-driven, with China recruiting Taiwanese citizens who are critical of their government to private group chats to spread manipulated photos and videos. China has significant influence on Taiwan’s information landscape, which can be manipulated during election cycles. However, it should be noted that Shen and Doublethink Lab are tied to the DPP, with Shen being the DPP’s mayoral candidate for Taipei in 2026. While this does not make Shen’s reporting untrue, it does speak to the partisan nature of efforts to address disinformation. With upcoming elections in 2026 and 2028, Taiwan must brace for renewed cognitive attacks. 

New Technology and Future Implications

Looking ahead to Taiwan’s local elections in 2026 and presidential election in 2028, it is crucial to understand China’s mission not just to manipulate narratives but also to build an information network. In August 2025, Vanderbilt University obtained leaked documents from the Chinese company GoLaxy, which scrapes data from social media to analyze public sentiment for corporations and the Chinese government. The documents revealed GoLaxy’s Smart Propaganda System (GoPro) technology, which collects information on specific Internet and social media users. GoLaxy claims that its GoPro technology tracks political situations to produce targeted content in real time. In December 2023, just one month before Taiwan’s presidential election, GoLaxy made recommendations regarding specific messages that they or other companies could push to exploit divisions in Taiwanese public opinion. The company also claims that it “has already produced certain political effects in relevant state departments.” According to GoLaxy, the end goal of GoPro technology is to bolster China’s voice and influence to “quickly make ‘the easterly wind overpowering the westerly wind’ a reality.” 

The GoLaxy documents reveal a significant shift in China’s information campaign strategy. The precision data monitoring and low-cost content creation enabled by AI technology have moved China’s focus away from broadcasting a single message across an entire domain and toward tailoring content to smaller groups and individuals. AI is facilitating a shift in disinformation campaigns; Taiwan does not just have to fact-check rumors and news articles, but instead must counter an entire AI-powered ecosystem of tailored disinformation. 

Chinese plans to influence Taiwan’s 2026 elections are already underway. Top CCP officials met in February during the Taiwan Work Conference and discussed establishing a task force to interfere in Taiwan’s upcoming local elections. The officials outlined using China’s united front to weaken “Taiwan independence forces,” a phrase that China often uses in reference to the DPP. Local elections are particularly vulnerable to outside interference as the fragmented information space gives China a broader scope to spread rumors. In 2026, there have already been two documented instances of Chinese information campaigns against Taiwan. In April, Taiwan’s National Security Bureau (NSB) detected 13,000 suspicious Internet accounts and 860,000 disputed messages. The accounts and messages flagged had posted AI-generated content related to foreign affairs, national defense, and the economy. The second wave of disinformation came following the Donald Trump–Xi Jinping summit in May. Taiwan’s NSB identified nearly 100 suspicious accounts and more than 9,000 disputed messages regarding the summit. These posts pushed China’s long-standing American Skepticism narrative, the idea that Taiwan cannot trust the United States as a partner and that only reunification will bring peace. Chinese cognitive warfare threatens not only Taiwan, but democracies worldwide. 

China’s AI Offensive on Taiwan: A Test Case for Democracy

AI technology has only improved since Taiwan’s last election in 2024. While Taiwan’s government, civil society organizations, and analysts are continually attempting to counter Chinese interference, doing so has become increasingly difficult. Chinese cognitive campaigns have evolved beyond state media narratives and into a web of disinformation and amplification that seeks to stoke divisions within Taiwan’s population. While Taiwan’s open society and partisan divides make it vulnerable to cognitive warfare, the island remains particularly resilient. Taiwan is a highly literate society, and China’s persistent attacks have heightened public awareness of disinformation campaigns. Taiwan’s advantages in AI expertise and technology production can also be harnessed in collaboration with other democratic societies, such as the United States, to create more resilient democracies across the globe. In the coming years, China’s cognitive warfare campaigns in Taiwan will serve as a case study either for how AI technology undermines democracy or for how governments can safeguard themselves from attack. What determines the outcome is whether Taiwan can draw on its technological advantages and alliances to protect its information environment. China’s cognitive warfare highlights both how AI technology can be used in new domains of conflict and how democracies can respond.

Featured Image: Unsplash | James Ang