China Likely Launched Large-Scale Cognitive Warfare Campaign Over Takaichi’s Taiwan Remark

The Yomiuri Shimbun

14:49 JST, March 23, 2026

China likely launched a major cognitive warfare campaign six days after Prime Minister Sanae Takaichi, speaking in November of last year, made a remark on a possible contingency involving Taiwan, according to analysis by The Yomiuri Shimbun and Sakana AI. The Yomiuri and the AI developer analyzed criticism of Japan by China on social media.

China seems to have decided how to respond to the situation while observing reactions in Japan.

This is the first time that cognitive warfare by China against Japan has been brought to light by analyzing a massive volume of social media posts using AI technology.

In cognitive warfare, people are influenced with fake information and narratives into which a speaker’s own viewpoints and assertions are mixed in, all to create a situation advantageous to one’s own country. Cognitive warfare is regarded as the sixth field of battle, after land, sea, air, space and cyberspace.

It was at a meeting of the House of Representatives’ Budget Committee on Nov. 7 that the prime minister suggested that if China were to impose a naval blockade on Taiwan, U.S. forces in the area may be attacked as well.

“If it involves the use of military force, it could by all means become a survival-threatening situation” for Japan, she said.

The Chinese government strongly opposed her remarks, and Japan-China relations deteriorated as China demanded its citizens to refrain from traveling to Japan and increased pressure by restricting exports of dual-use products.

The Yomiuri Shimbun and Sakana AI analyzed about 400,000 posts criticizing Japan on X and Chinese social media platform Weibo, published from late October last year to January this year. Those posts by major accounts affiliated with the Communist Party of China, such as the accounts of Chinese government organizations and government-affiliated media, were then extracted out. The content of those posts was analyzed, including their nuances, using a large language model, or LLM.

Few posts attacking Japan were found from Nov. 7 to 9. They briefly grew in number on Nov. 10, the day China’s Foreign Ministry criticized Takaichi’s remark, but there were fewer again on Nov. 11 and 12, before a sudden surge from Nov. 13. China did not immediately react against the remarks. After six days of ‘silence,’ the Chinese government summoned Japanese Ambassador Kenji Kanasugi in Beijing on Nov. 13, and this is understood to have been the signal for the full-scale start of cognitive warfare. Views of posts that criticized Japan across X increased sharply from Nov. 14.

Posts criticizing China also increased on X from Nov. 8 to 12. Most of these posts expressed anger at Xue Jian, the Chinese consul general in Osaka, who reacted to Takaichi’s remark by saying in an X post, “There would be no choice but to cut off that filthy head without a moment’s hesitation.”

The AI model used for the analysis found that during the six days of silence, China implemented its cognitive warfare strategy in three stages: “consideration,” “initial public signal” and “full-scale launch.” The AI model claims that when the strategy against Japan was drawn up, the strong reactions from Japan were also taken into consideration. An analysis of Weibo accounts affiliated with China’s Communist Party found the same pattern of posting.

“Decision making in China is like a black box, but I don’t find the analysis strange,” said a person close to the Japanese government.

“It seems [Chinese] Foreign Minister Wang Yi and others initially took a wait-and-see stance on the prime minister’s remark,” said another person close to the government, who assumed that those in the Chinese government had to report to Chinese President Xi Jinping about Takaichi’s remark amid increasing criticism against China.