Senior National Security Columnist, The Cipher Brief
OPINION — Taiwan’s Presidential election last Saturday took place amid widespread concerns that China would use Artificial Intelligence (AI) driven disinformation to affect the outcome.
Apparently, it didn’t. The election came off close as generally expected with Taiwan’s current vice president, Lai Ching-te, and his ruling Democratic Progressive Party (DPP) declaring victory, while his two opposition rivals both conceded defeat. According to Taiwan’s Central Election Commission, Lai won with just over 40% of the total votes, Kuomintang (KMT) party candidate Hou Yu-ih got 33.49% of the votes, with Taiwan People’s Party (TPP) candidate Ko Wen-je getting 26.45%.
In a December 19, 2023, article, Taiwan’s Election: 2024’s Canary in the Coal Mine for Disinformation against Democracy, Dylan Welch, China Technology Analyst at the German Marshall Fund’s Alliance for Securing Democracy, wrote that Taiwan was experiencing “a surge in AI-generated disinformation, including images, videos, audio, and text, all aimed at influencing electoral outcomes. The island not only finds itself at the epicenter of this new wave of tech-enabled electoral meddling but is also pioneering responses to navigate this unprecedented challenge in digital democracy.”
With the U.S. facing its own Presidential election this November, I thought it worthwhile to look into the Taiwan experience after reading in Welch’s article that Taiwan had adopted “pioneering responses” to the AI-driven digital election threat.
Today’s constant barrage of information makes it easy for countries to wage disinformation campaigns and your emotions are the weapon of choice. Learn how disinformation works and how we can fight it in this short video. This is one link you can feel good about sharing.
As Welch put it, “Because of the relentless barrage of disinformation, the country’s model for countering disinformation is one of the world’s most developed.”
Although Taiwan had been a major target for Chinese misinformation for years, a turning point was reached in 2018, after Su Chii-Cherng, a Taiwanese diplomat in Osaka, Japan, committed suicide after Chinese media outlets distributed a fake story claiming that he failed to help Taiwanese citizens escape during a typhoon there.
The fake story, posted on major Taiwan social media sites, was that Chinese diplomats in Japan had sent buses to rescue stranded Taiwanese tourists at Osaka’s Kansai International Airport, but the tourists only could board if they identified as “Chinese.” That led to a Taiwanese public outcry against Su, the Taiwan diplomat on scene — and his assumed incompetence — which led to his committing suicide.
In 2019, Taiwan’s legislature passed the 2019 Anti-Infiltration Law that blocked foreign entities from spreading misinformation or otherwise interfering in Taiwan’s elections. The act did not mention China, but its target is Chinese actors and Taiwanese citizens with connections to China. In 2020, after the DPP government objected to continued disinformation spread on Chinese media platforms, Taiwan banned several of them from the Taiwanese market.
Since 2020, every Taiwanese ministry has established a team to detect disinformation campaigns and rapidly respond with a counter-narrative. The Department of Cyber Security protects websites and databases from hackers.
As of December 2023, Welch wrote, “There have been fines leveled at television stations for broadcasting information that was found to be false. There have [also] been concerns that the law will impede freedom of speech in Taiwan.”
Taiwan has also developed an ecosystem of independent civil organizations working on disinformation. Doublethink Lab, which was founded in 2019 and describes itself as “researching malign Chinese influence operations,” has a monitoring hub using AI and human arbiters to detect Chinese information operations before they reach the mainstream.
Subscriber+Members have a higher level of access to Cipher Brief Expert Perspectives and get exclusive access to The Dead Drop, the best national security gossip publication, if we do say so ourselves. Find out what you’re missing. Upgrade your access to Subscriber+ now.
CoFacts calls itself a Taiwan “information checking platform operated through crowd collaboration and chatbot to have discrete messages of unknown credibility carefully reviewed and discussed through the joint efforts of the public.” According to Al Jazeera, “Cofacts automatically responds to fake or misleading messages circulated on the LINE [Taiwan-based] messaging app with a sourced report. Fact checks are written and reviewed by a group of more than 2,000 volunteers, including teachers, doctors, students, engineers and retirees – anyone who wants to be a fact-checker can become one.”
The Taiwan Fact Check Center is another civic society group whose chatbot onLINE social media – a version of Facebook — allows users to check for disinformation or misinformation against a database that is updated by its team of fact checkers.
Taiwan FactCheck Center CEO, Eve Chiu, explained that her group monitors different social media platforms and exposes false information items that may have been presented in the mainstream media as “Breaking News” or “Exclusives.” but which were actually taken from faked social media or misquoted government documents.
For example, last August, when Taiwan’s Vice President Lai, who at the time was a presidential candidate, visited Paraguay, a document with his signature appeared that apparently gave Paraguay a large amount of money for its social housing. It created a negative uproar in Taiwan which was having its own debate about social housing. Chiu’s fact checkers found the document was a fake, created to discredit Lai. In reality, it was a copied Paraguayan government paper dated 2018, but signed by Taiwan’s then-President Tsai Ing-wen.
Chiiu described it as, “Almost the same. Even the location of the stamp, even the signature – right hand signature is the same. This has been [a remade] document. And it’s [a] very sophisticated way to do this fake document and post it.”
In November, the Taiwan FactCheck Center flagged a video posted online that showed DPP’s front-runner Lai, oddly claiming that the opposition represents the majority view in Taiwan. However, the FactCheck Center pointed out the possibility of AI manipulating his voice because Lai’s mouth appeared to move in an unnatural way. Taiwanese authorities later confirmed the video as a deepfake, according to later news reports.
When rumors developed on line that DPP’s vice-presidential candidate Hsiao Bi-khim is a U.S. citizen, FactCheck Center publicized that while Hsiao once held U.S. citizenship, she had renounced it in 2002.
Auntie Meiyu is said to be the most widely used chatbot app for Taiwanese private chat groups. In 2018, it was first developed to deal with disinformation sent by family elders in a family chat group. Since 2020, it has used advanced AI technology to prevent communication fraud.
As its website says, “Add Auntie Meiyu into your group chat! She will silently lurk in the chat room and never interrupts your conversation. Auntie Meiyu will speak up only when she notices some questionable information being posted. She then provides verified information and helps you identify fake news.”
You can, according to its website, “Ask Auntie Meiyu directly about any specific news or articles with a private message. If there are no related news or sources, you can also contribute by reporting the rumor along with any relevant information.”
Taiwan’s trust-tech company Gogolook, in 2022, was recommended for users by the Central Election Commission. Gogolook also said Auntie Meiyu had been adopted by over 520,000 users; and that Auntie Meiyu verified/debunked suspicious information 1.67 million times in 2022.
Not all deepfake videos were created with the intent to affect elections, although they may have appeared that way.
Looking for a way to get ahead of the week in cyber and tech? Sign up for the Cyber Initiatives Group Sunday newsletter to quickly get up to speed on the biggest cyber and tech headlines and be ready for the week ahead. Sign up today.
Last November, Taiwan’s Criminal Investigation Bureau (CIB) warned about fraudulent crypto currency advertisements circulating online that featured a video clip of Taiwan President Tsai and Vice President Lai discussing a crypto investment. Using technology, the video mimicked the voices of Tsai and Lai and adjusted original footage to make their lips match their words. The Chinese-language ads had clickable links promising easy money, according to the CIB.
Coming in the midst of Taiwan’s Presidential election campaign, some attributed it to the Chinese attempting to influence voting, but the CIB described it as a scam using the political leaders to lure investors.
Welch wrote that “Chinese influence operations are strategically aimed at swaying Taiwan’s election to favor of an agenda of so-called peaceful reunification, which would lead to support of the KMT Party.”
On December 21, Taiwanese authorities arrested online journalist Lin Hsien-yuan, whose poll showed for the first time, that KMT’s candidate Hou leading to win the presidential election. The poll results briefly went viral in Taiwan, according to Radio Free Asia. Lin had pretended to have interviewed or polled 300 people. Taiwanese prosecutors, using the Anti-Infiltration Act, said Lin’s findings were faked and may have been encouraged by Chinese Communist Party officials from across the Taiwan Strait.
Taiwan, population some 24 million, for its January 2024 election built up governmental, media and private defenses against misinformation, fake news videos, false AI-generated hosts and voice-overs that circulated on YouTube, Instagram, Facebook, TikToc, and X.
With a population of 336 million, can the U.S. do anything similar via government, media and private groups, before the November 2024 election?
The Cipher Brief is committed to publishing a range of perspectives on national security issues submitted by deeply experienced national security professionals.
Opinions expressed are those of the author and do not represent the views or opinions of The Cipher Brief.
Have a perspective to share based on your experience in the national security field? Send it to Editor@thecipherbrief.com for publication consideration.
Read more expert-driven national security insights, perspective and analysis in The Cipher Brief
Pulitzer Prize Winning Journalist Walter Pincus is a contributing senior national security columnist for The Cipher Brief. He spent forty years at The Washington Post, writing on topics that ranged from nuclear weapons to politics. He is the author of Blown to Hell: America’s Deadly Betrayal of the Marshall Islanders.