TAIPEI (Taiwan News) — The National Security Bureau on Sunday released a report warning that China is ramping up cognitive warfare using tech firms, marketing networks, and AI tools to spread disinformation and manipulate public opinion.
The bureau said the CCP continues to wage cognitive warfare against Taiwan by amplifying major current events to achieve strategic objectives, per CNA. These goals include stoking internal divisions in Taiwan, weakening the public will to resist China, undermining allies’ support for Taiwan, and promoting acceptance of Beijing’s position.
The bureau said the report aims to help the public better understand the harm posed by CCP influence operations to democratic societies. Once narratives and messaging tones are set, China’s party, state, and military systems mobilize IT companies, marketing firms, and online influence groups to deploy five main disinformation tactics.
China’s Cyberspace Administration, Ministry of State Security, and the PLA’s Political Work Department work with tech firms such as Golaxy, Meiya Pico, and Warming High-Tech to use web crawlers. These tools collect data on Taiwanese politicians, elected officials, opinion leaders, and election-related polling and social media metrics to enable targeted propaganda and attacks.
China’s Central Publicity Department and Ministry of Public Security coordinate with marketing firms such as Haixunshe, Haimai, and Huya to create websites masquerading as international media outlets. Websites such as “Aisa Korea” and “Austria Weekly” collaborate to spread official CCP narratives.
Beijing has backed the Chinese marketing firm Wubianjie Group to operate Facebook-based content farms that post sensational material to drive traffic. The firm also runs accounts on platforms such as Threads and X that begin with soft content before shifting to political messaging to influence Taiwanese users.
China’s Ministry of Public Security has used the Dragonbridge online influence group to conduct operations in more than 20 languages across over 180 platforms worldwide, the bureau said. It added that the Cyberspace Administration of China, the United Front Work Department, and the PLA’s Cyberspace Force have directed firms including Zhongkedianji, Beijing Xingguang, and OneSight to build databases and automated systems controlling large numbers of bot accounts.
Companies such as China North Industries Group are developing AI systems that combine opinion data collection, automated video generation, and precision targeting. Firms including Magic Data and iFlytek have been commissioned to place ads on recruitment websites to entice Taiwanese users to record their voices, which could then be used to mimic local accents, increasing the credibility of AI-generated content.
During China’s military exercises targeting Taiwan in April 2025, more than a dozen PTT user accounts were hijacked. Hackers used compromised IoT devices and rented foreign servers to spread false claims, including that China had blockaded Taiwan’s natural gas shipments and that Chinese warships had entered Taiwan’s 24-nautical-mile contiguous zone.
The bureau said intelligence agencies identified more than 45,000 inauthentic accounts last year and collected over 2.314 million pieces of disinformation. More than 3,200 pieces of disinformation were reported to government agencies to enable timely countermeasures.
The bureau said government bodies and think tanks in the US, EU, Australia, France, and other countries have also warned about CCP information manipulation. It said Taiwan held more than 80 security dialogues and intelligence conferences with international partners last year to strengthen democratic cooperation against Beijing’s cognitive warfare.
