Suspicious videos that began circulating in Taiwan this month seemed to show the country’s leader advertising cryptocurrency investments.
President Tsai Ing-wen, who has repeatedly risked Beijing’s ire by asserting her island’s autonomy, appeared to claim in the clips that the government helped develop investment software for digital currencies, using a term that is common in China but rarely used in Taiwan. Her mouth appeared blurry and her voice unfamiliar, leading Taiwan’s Criminal Investigation Bureau to deem the video to be almost certainly a deepfake — an artificially generated spoof — and potentially one created by Chinese agents.
For years, China has pummeled the Taiwanese information ecosystem with inaccurate narratives and conspiracy theories, seeking to undermine its democracy and divide its people in an effort to assert control over its neighbor. Now, as fears over Beijing’s growing aggression mount, a new wave of disinformation is heading across the strait separating Taiwan from the mainland before the pivotal election in January.
Perhaps as much as any other place, however, the tiny island is ready for the disinformation onslaught.
Taiwan has built a resilience to foreign meddling that could serve as a model to the dozens of other democracies holding votes in 2024. Its defenses include one of the world’s most mature communities of fact checkers, government investments, international media literacy partnerships and, after years of warnings about Chinese intrusion, a public sense of skepticism.
The challenge now is sustaining the effort.
“That is the main battlefield: The fear, uncertainty, doubt is designed to keep us up at night so we don’t respond to novel threats with novel defenses,” said Audrey Tang, Taiwan’s inaugural digital minister, who works on strengthening cybersecurity defenses against threats like disinformation. “The main idea here is just to stay agile.”
Taiwan, a highly online society, has repeatedly been found to be the top target in the world for disinformation from foreign governments, according to the Digital Society Project, a research initiative exploring the internet and politics. China was accused of spreading rumors during the pandemic about the Taiwanese government’s handling of Covid-19, researchers said. Representative Nancy Pelosi’s visit to the island as speaker of the House last year set off a series of high-profile cyberattacks, as well as a surge of debunked online messages and images that fact checkers linked to China.
For all of Beijing’s efforts, however, it has struggled to sway public opinion.
In recent years, Taiwan’s voters have chosen a president, Ms. Tsai, from the Democratic Progressive Party, which the Communist Party views as an obstacle to its goal of unification. Experts and local fact checkers said Chinese disinformation campaigns were a major concern in local elections in 2018; the efforts seemed less effective in 2020, when Ms. Tsai recaptured the presidency in a landslide. Her vice president, Lai Ching-te, has maintained a polling lead in the race to succeed her.
Ms. Tsai has repeatedly addressed her government’s push to combat Beijing’s disinformation campaign, as well as criticism that her strategy aims to stifle speech from political opponents. At a defense conference this month, she said: “We let the public have knowledge and tools that refute and report false or misleading information, and maintain a cautious balance between maintaining information freely and refusing information manipulation.”
Many Taiwanese have developed internal “warning bells” for suspicious narratives, said Melody Hsieh, who co-founded Fake News Cleaner, a group focused on information literacy education. Her group has 22 lecturers and 160 volunteers teaching anti-disinformation tactics at universities, temples, fishing villages and elsewhere in Taiwan, sometimes using gifts like handmade soap to motivate participants.
The group is part of a robust collective of similar Taiwanese operations. There is Cofacts, whose fact-checking service is integrated into a popular social media app called Line. Doublethink Lab was directed until this month by Puma Shen, a professor who testified this year before the U.S.-China Economic and Security Review Commission, an independent agency of the U.S. government. MyGoPen is named after a homophone in the Taiwanese dialect for “don’t fool me again.”
Citizens have sought out fact-checking help, such as when a recent uproar over imported eggs raised questions about videos showing black and green yolks, Ms. Hsieh said. Such demand would have been unthinkable in 2018, when the heated emotions and damaging rumors around a contentious referendum inspired the founders of Fake News Cleaner.
“Now, everyone will stop and think: ‘This seems odd. Can you help me check this? We suspect something,’” Ms. Hsieh said. “This, I think, is an improvement.”
Still, fact-checking in Taiwan remains complicated. False claims swirled recently around Mr. Lai, an outspoken critic of Beijing, and his visit to Paraguay this summer. Fact checkers found that a memo at the center of one claim had been manipulated, with changed dates and dollar figures. Another claim originated on an English-language forum before a new X account quoted it in Mandarin in a post that was shared by a news website in Hong Kong and boosted on Facebook by a Taiwanese politician.
China’s disinformation work has had “measurable effects,” including “worsening Taiwanese political and social polarization and widening perceived generational divides,” according to research from the RAND Corporation. Concerns about election-related fake news drove the Taiwanese government last month to set up a dedicated task force.
Taiwan “has historically been Beijing’s testing ground for information warfare,” with China using social media to interfere in Taiwanese politics since at least 2016, according to RAND. In August, Meta took down a Chinese influence campaign that it described as the largest such operation to date, with 7,704 Facebook accounts and hundreds of others across other social media platforms targeting Taiwan and other regions.
Beijing’s disinformation strategy continues to shift. Fact checkers noted that Chinese agents were no longer distracted by pro-democracy demonstrations in Hong Kong, as they were during the last presidential election in Taiwan. Now, they have access to artificial intelligence that can generate images, audio and video — “potentially a dream come true for Chinese propagandists,” said Nathan Beauchamp-Mustafaga, a RAND researcher.
A few months ago, an audio file that seemed to feature a rival politician criticizing Mr. Lai circulated in Taiwan. The clip was almost certainly a deepfake, according to Taiwan’s Ministry of Justice and the A.I.-detection company Reality Defender.
Chinese disinformation posts appear increasingly subtle and organic, rather than flooding the zone with obvious pro-Beijing messages, researchers said. Some false narratives are created by Chinese-controlled content farms, then spread by agents, bots or unwitting social media users, researchers say. China has also tried to buy established Taiwanese social media accounts and may have paid Taiwanese influencers to promote pro-Beijing narratives, according to RAND.
Disinformation that directly addressed relations between China and Taiwan grew rarer from 2020 to 2022, the Taiwan Fact Check Center said last month. Instead, Chinese agents seemed to focus more on stoking social division within Taiwan by spreading lies about local services and health issues. Sometimes, other experts said, questionable posts about medical remedies and celebrity gossip guided viewers to conspiracy theories about Taiwanese politics.
The ever-present menace, which the Taiwanese government calls “cognitive warfare,” has led to several aggressive attempts at a crackdown. One unsuccessful proposal last year, modeled after regulations in Europe, would have imposed labeling and transparency requirements on social media platforms and forced them to comply with court-ordered content removal requests.
Critics denounced the government’s anti-disinformation campaign as a political witch hunt, raising the specter of the island’s not-so-distant authoritarian past. Some have pointed out that Taiwan’s media ecosystem, with its diverse political leanings, often produces pro-Beijing content that can be misattributed to Chinese manipulation.
At an event in June, President Tsai stressed that “well-funded, large-scale disinformation campaigns” were “one of the most difficult challenges,” pitting Taiwanese citizens against one another and corroding trust in democratic institutions. Disinformation defense, she said, must be “a whole-of-society effort.”
Fact checkers and watchdog groups said public apathy was a concern — research suggests that Taiwanese people make limited use of fact-checking resources in past elections — as was the risk of being spread too thin.
“There’s mountains of disinformation,” said Eve Chiu, the chief executive of the Taiwan FactCheck Center, which has around 10 fact checkers working each day. “We can’t do it all.”
Attempts to increase interest in media literacy have included a nationwide campaign, “humor over rumor,” which leveraged jokey meme culture and a cute dog character to debunk false narratives. In September, the Taiwan FactCheck Center also held a national virtual competition for youths that drew students like Lee Tzu-ying, Cheng Hsu-yu and Lu Hong-yu.
The three civics classmates, who finished in third place, acknowledged that Taiwan’s raucous politics allowed disinformation to breed confusion and chaos. Their Taiwanese peers, however, have learned caution.
“If you see something new, but don’t know if it is true or false, you need to verify it,” Ms. Lee, 16, said. “I just want to know the truth — that’s very important to me.”