Taiwan is the country most affected by disinformation in the world, thanks to Chinese efforts. Taiwan is presumably an easy target due to its free information environment, high proportion of citizens who are online and cultural and linguistic proximity to China. Although Taiwan has been dealing with different forms of Chinese information interference for decades, China’s efforts became significantly more convincing once it had a compelling narrative — the China success story — and the internet. Chinese disinformation efforts in Taiwan significantly ramped up after 2016, when then-President Tsai Ing-wen (蔡英文) was elected and the world was alerted to Russia’s manipulation of the online information environment. China, following Russia’s playbook, uses Taiwan as a test bed for its online disinformation campaigns.
China’s disinformation campaigns aim to undermine confidence in Taiwan’s democratic governance and exacerbate political divisions. Yet, Taiwan’s democratic pluralism and the inherent weaknesses of authoritarian messaging limit the effectiveness of disinformation efforts. While generative AI is making Chinese disinformation more sophisticated, Taiwan is adapting through community-driven solutions that aim to foster trust and empower citizens to detect and counter falsehoods.
“Given that Taiwan is such a small country, and … on the front line of the China threat … we all have to work together” to combat disinformation, said Chiaoning Su (蘇巧寧), a professor at Oakland University who is an expert on the intersection of journalism and democracy.
Chinese Disinformation in Taiwan
Chinese disinformation is spread by various agents and takes many forms. These agents include the 50 Cent Army (五毛黨) — a group of internet commentators who are employed by the Chinese government to spread propaganda — and state-affiliated content farms. They spread messages that range from “soft” topics that promote China’s attractiveness to “hard” topics such as messages that praise communist governance or claim that democracy is chaotic and inefficient. With the advent of AI, disinformation may also take the form of misleading images and videos such as deep fakes.
Two main characteristics of Taiwan’s information environment shape Taiwan’s vulnerability — and China’s approach — to information manipulation. First is high levels of political polarization, which allows China to exacerbate existing divisions in Taiwanese society, such as political party affiliation and trust toward democratic institutions. Polarization also leads Taiwanese to “politicize everything,” said Su, making it harder for people to process information in an objective way. A second factor is low levels of trust in traditional media. “When you don’t trust the public institution, then you give more weight [to] interpersonal communication … You tend to believe your neighbor, your friends, your boyfriend, girlfriend,” said Su. This lends more credibility to social media platforms like Line and Facebook as a source of information. In fact, the majority of Taiwanese use social media for the news. As a result, disinformation makes its largest impact via closed communication apps like Line — group chats, essentially — said Su.
Taiwan’s Intrinsic Resilience to China’s Evolving Disinformation Strategies
Polarization, perhaps ironically, is also the first line of defense when it comes to Taiwan’s information environment. “The beauty of a divided society is that … only half of the population will buy the story,” said Yi-Suo Tzeng (曾怡碩), a research fellow at the institute for national defense and security research specializing in information operations and cognitive warfare. This is an inherent advantage of democracy, as polarization is made possible by democratic freedoms that allow for a plurality of political opinions.
Intrinsic pitfalls of authoritarianism also mitigate the impact of disinformation. First is the issue of Chinese messaging itself. Despite shared history and language, China’s messaging often feels “alien” to Taiwanese, for example the use of simplified Chinese instead of traditional, said Tzeng. Second, Chinese disinformation also tends to reuse certain terms and themes, such as U.S. skepticism, that make it easy to flag as propaganda. Bureaucrats are more focused on pleasing Chinese President Xi Jinping (習近平) and demonstrating ideological correctness than persuading Taiwanese. “I don’t think they want to understand Taiwan anymore … I think they’ve given up,” said Tzeng.
However, some of this natural resilience might be eroding with the proliferation of generative AI, which is making it easier for China to mimic Taiwanese style and create new, more nuanced narratives. The phrase “you are what you eat” applies to AI, meaning that if fed culturally specific Taiwanese data, AI could create narratives that sound increasingly natural to Taiwanese internet users, said Tzeng. China could also conceivably use AI to constantly track current events to come up with new narratives so that it doesn’t have to rely on overdone themes like U.S. skepticism. Taiwan doesn’t expect that it can prevent disinformation or weather it unscathed. “No matter how you try to mitigate it, it will keep coming,” said Tzeng. This is where grassroots fact checking initiatives are of great use.
Civic Solutions and Community-Driven Fact Checking
Taiwan’s vibrant fact-checking ecosystem includes fact-checking organizations like Cofacts, Taiwan FactCheck Center and MyGoPen, as well as organizations like the Taiwan Information Environment Research Center, which provides literacy education and conducts multidisciplinary scientific research on information manipulation. Taiwan AI Labs uses AI technology to track the flow of disinformation in Taiwan. A unique aspect of this ecosystem is that these organizations are very willing to work together, as “the danger, the threat that we are facing, is real,” said Su from Oakland University. Freedom House credits Taiwan’s diverse and cohesive civil society responses and media literacy efforts with making Taiwan “very resilient” to China’s persistent disinformation campaigns.
One of these organizations, Cofacts, was developed by members of the Taiwanese civic hacking community g0v in 2016 to break echo chambers in Taiwan by showing people what parts of what they are reading or seeing are misinformation, opinion or bias.
Cofacts embraces the Taiwanese tendency to distrust official sources of information — whether that be from the government or traditional media — said Billion Lee (李比鄰), co-founder of Cofacts. “Since [Taiwanese people] receive information from their friends, we want to become their friends, and we imitate the way they communicate,” Lee said. Line users can add Cofacts as a contact and forward information to it like they would to their friends. The chatbot will tell them whether the information is accurate, the reason why and provide references. If the user sends the chatbot unverified information, the chatbot uses generative AI to give the user advice on how to factcheck the information themselves. Over 300,000 people use Cofacts, sending over 6,000 inquiries to the chatbot each week.
Collaboration is an anchoring value of Cofacts, said Lee. Its crowdsourced database is continuously updated with verified information by nearly 3,000 volunteer fact checkers. As the first fact checking chatbot, Cofacts provides their code as open source, so that any individual or organization can make their own version of the chatbot. Lee also embraces this ethic of collaboration within Cofacts’ marketing strategy, going to churches, club meetings and traveling all over Taiwan, including into the mountains to meet with indigenous groups. Beyond introducing the chatbot to new communities, Cofacts improves its system by talking to people and figuring out what is important to them, said Lee.
If the threat of war exists, “Taiwan cannot … just wait for the U.S. to help,” said Lee. Disinformation is a weapon, and we can respond by providing openness and transparency to build trust at the grassroots level.
Leave a Reply