Election Interference and Information Manipulation

Democracies worldwide can learn from Taiwan’s response to external intervention in its recent elections.
June 06, 2024
11 min read
Photo credit: Skorzewiak / Shutterstock.com
2024 is a year of elections. As countries holding them prepare, they are also bracing for foreign interference.

The West remains focused on Russia as a sponsor of electoral disinformation and manipulation, but in recent years the People’s Republic of China (PRC) has emerged as another key actor to watch.

Taiwan has been a key target for the PRC in this regard for many years. It was no different in January, when Taiwan’s voters elected, for a historic third consecutive time, the Democratic Progressive Party (DPP) candidate for president. The experience, however, offers valuable lessons for other democracies confronting Beijing’s nefarious strategies.

Against this background, GMF’s Indo-Pacific program convened a closed-door roundtable in Washington, DC on April 10 and offered Taiwanese NGOs a platform to present their analysis of the recent election interference to officials, policymakers, and think-tank representatives from around the world. Participating Taiwanese NGOs included the Taiwan FactCheck CenterCoFacts, the Taiwan Information Environment Research Center (IORG), and Doublethink Lab. A Taiwanese government official also gave a presentation.

The report below presents key findings from the roundtable, with additional analysis from the author and the participating NGOs.

Beijing’s Extensive Toolkit

The PRC’s efforts to influence Taiwan’s January election employed techniques that combined military, economic, and diplomatic action with information operations. These efforts included holding provocative military exercises near Taiwan's maritime borders, threatening people perceived as “pro-independence”, offering economic incentives to Taiwanese businesses, and trying to co-opt other social groups that support closer ties with Beijing.

The PRC’s overt and covert disinformation efforts were aimed at manipulating public perceptions in Taiwan and destabilizing society by disseminating false or misleading narratives through social media platforms, state-controlled media outlets, proxy accounts, and Beijing-aligned voices in Taiwan. Importantly, a given action often appeared coordinated to reinforce another. Military maneuvers, for example, accompanied disinformation campaigns hinting at imminent conflict. The objective was to spread fear.

PRC tactics targeting Taiwan blur the line between acceptable influence and illegitimate interference, but they often lean toward the latter, especially when considering their cumulative effect. Beijing seeks to exacerbate social divisions and undermine trust in democratic institutions through this approach. The ultimate goal is to erode Taiwan’s ability to resist PRC pressure and create the impression that unification with China is the only viable option.

Throwing Mud

The PRC's election interference specifically aimed to undermine the DPP, presenting it as a party intent on destroying Taiwan, while boosting the Kuomintang (KMT) and other Beijing-friendly opposition groups. But the PRC’s activities also sought to discredit Taiwanese democracy by presenting it as chaotic and illegitimate.

Information manipulation, including outright disinformation such as the dissemination of conspiracy theories and demonstrably forged documents, played a key role before, during, and after the election. Western observers have ascribed even more-destructive tactics to create chaos and erode trust to Russia's playbook, but the PRC has long employed similar strategies in Taiwan. Taiwanese NGOs have been particularly strong in documenting these cases and providing fact-checking services and resources to civil society to stay ahead of barrages of false and misleading information.

A key PRC tactic has been to propagate narratives through official channels and proxies that promote skepticism of the United States. IORG has identified 84 such narratives used in Taiwan. They include questioning US reliability and its commitment to Taiwan in case of attack. Another narrative is collusion between the United States and Taiwan’s elites, which will lead to war and Taiwan’s destruction. 

Overall, researchers found escalating themes over time about US abandonment of Taiwan. Some narratives promoted the idea that the Washington is preparing for Taiwan’s destruction by extracting all its useful resources. Part of this alleged plot was the request to Taiwanese semi-conductor manufacturer TSMC to build a plant in Arizona.

The campaign to undermine the DPP painted the party as a source of regional conflict and instability that would inevitably lead to confrontation. KMT candidates adopted this perspective even though Taiwanese NGOs documented many pieces of false and misleading news about leading DPP politicians, including Tsai Ing-wen, Lai Ching-te, and Bi-khim Hsiao

Narratives and messages attacking election integrity were also common before, during, and after the vote. Some of these efforts, although unsuccessful in eroding confidence more broadly, suggested that poll workers stuffed ballot boxes or that voting devices used disappearing ink that would invalidate ballots. Prior elections saw only small disinformation campaigns alleging fraud, but their volume surged around January’s vote. Video clips on YouTube and TikTok purported to give eyewitness accounts of election tampering, and some videos attracted millions of views

Other information manipulation campaigns successfully played on social anxieties and spilled over into protests. One campaign exploited racist sentiment after Taipei signed a memorandum with New Delhi to bring 100,000 Indian migrant workers to Taiwan. A small in-person demonstration occurred in December 2023, which was highlighted by Doublethink Lab and others for its ability to turn online protests into offline action, still considered a novelty in the Taiwanese context.

Flooding the Space

PRC information manipulation and interference strategies rely in part on unofficial and local voices to make campaigns more credible and effective. The hiring of influencers, whether Chinese or not, has been repeatedly documented in many countries. Taiwan is no exception, especially when it comes to TikTok.

Beijing has long pursued "public opinion management" tools to make coordinated inauthentic behavior on social media more effective and create the impression of widespread support for pro-PRC positions. Public procurement documents show that PRC party and government departments also work with commercial companies to do this. The interplay between official accounts and bot networks is well known. A 2021 Oxford Internet Institute study suggests that such networks are used to amplify content from official PRC embassy accounts on X (formerly Twitter).

IORG researchers have analyzed cross-platform datasets for identical or near-identical content to identify suspiciously synchronized social media activity that mirrors PRC state-owned media content and narratives. They have found likely PRC proxies on TikTok that circulated content nearly identical to that on state media accounts on the PRC’s version of the platform, Douyin, which is also owned by ByteDance.

Short clips were a primary disinformation vehicle around the election in January. TikTok stood out as a key platform for amplifying PRC narratives and appears to have had at least some impact on the perceptions of the Taiwanese public. An IORG poll showed that the platform’s users are more likely than others to agree with pro-Beijing narratives or false narratives originating from or spread by PRC accounts. More than half of Taiwan’s TikTok users believed that the Tsai administration’s pro-US position provoked the PRC while only 38% of non-users shared that opinion.

Accounts that could not be directly linked to official PRC sources were often the sources of the most destructive aspects of information manipulation before and during the election. Doublethink Lab found that, compared to Taiwan’s 2022 election, the largest information manipulation campaigns around January’s vote relied more on inauthentic social media amplification through unofficial accounts. A common pattern saw official PRC voices amplifying disinformation posted by anonymous accounts or pro-KMT Taiwanese media. This was likely the case because non-official accounts are more effective and official PRC actors prefer to distance themselves from certain pieces of obvious disinformation to maintain plausible deniability.

More Generative AI

From audio and video manipulation to text generation, generative AI played a bigger role this year than in previous elections. Fake clips often featured political figures or other prominent individuals. One AI-generated video circulated ahead of January’s vote showed a US congressman announcing accelerated weapons sales to Taiwan in case of a DPP win. Deepfake videos promoted an AI-generated e-book, “The Secret History of Tsai Ing-wen”, which was subsequently featured in inauthentic social media accounts.

“Cheapfakes”, media altered through conventional technologies or simply presented out of context, were even more common than deepfakes, according to the Taiwan FactCheck Center. Such deceptions included photoshopped official documents and misleading video footage. Though less sophisticated than deepfakes, cheapfakes can effectively spread false narratives on social media, and malicious actors in many cases may favor them because they are easy to create and disseminate.

As generative AI improves, it will be increasingly used in information manipulation campaigns. There is an urgent need to invest in research to develop methods for easier and, ideally, automated detection of AI-created media.

A United Front Playbook

Not all pro-PRC voices in Taiwanese information spaces are inauthentic accounts or paid influencers. Beijing has also tried to exploit Taiwanese voices as part of the Chinese Communist Party’s (CCP) classic “United Front” approach that aims to sway pro-PRC groups and isolate opponents labeled as separatists. In Taiwan, much of this effort targets businesspeople, especially those who support the KMT. Local media have also reported that hundreds of officials have received discounted trips to the PRC, allegedly funded by Beijing’s Taiwan Affairs Office. Some Taiwanese authorities have opened legal proceedings against participants of those trips. 

The United Front outreach into Taiwanese business circles was highlighted by CCP official Wang Huning's address to Taiwanese CEOs at a cross-strait business summit in Nanjing in November 2023. He assured attendees of opportunities to participate in the PRC's development. At the same time, Beijing used economic penalties to undermine the DPP by, for example, ending in December 2023 tariff cuts on 12 chemicals imported from Taiwan. Beijing blamed the party's stance on independence for hindering another outcome. And just three days before the January election, Beijing announced plans to terminate additional tariff cuts, with Chinese state media openly blaming the DPP for that decision, too. Lastly, Beijing spread messaging about economic retaliation in case of a DPP win, which inauthentic social media amplified.

An Example To Follow

The Taiwanese case illustrates an intense election interference campaign by the PRC, whose actions are particularly instructive for other democracies because they reveal the entirety of Beijing’s toolkit. Taiwan’s proactive measures in response can serve as best practices for other democracies in similar circumstances. Moreover, Taiwan’s civil society and government have meticulously documented the PRC’s most recent campaign, which provides a trove of data for analysis and evaluation. This can help craft more effective countermeasures.

Despite their sustained efforts, the PRC campaigns proved largely unsuccessful in preventing the election of DPP candidates (though it may have succeeded in influencing public perceptions in Taiwan on other issues). The overall failure can be attributed partly to prevailing public opinion and the KMT’s limited popularity, which Beijing failed to change. However, successful monitoring, early detection, and countermeasures by Taiwanese authorities and civil society likely also contributed to preventing the campaigns from making deeper inroads.

This outcome provides insights for other democracies into safeguarding their own institutions, especially as the threat of PRC interference grows. Many lessons may also prove valuable for countering interference from other malicious actors.

Beijing, however, is unlikely to replicate exactly its playbook in other elections. Taiwan is a "core interest", which likely permits more resources to be dedicated to meddling in its affairs. The PRC’s clearly partisan goal of pushing the KMT is likely also different from its designs in other countries. Though allegations of partisan interference have arisen in some countries, including Canada, the PRC still does not want to be perceived as supporting any particular candidate in most elections. Instead, it is likely to focus on its broader aim to discredit democracy and showcase it as chaotic and declining. All democracies should beware that. 

Integrating Civil Society

Taiwan can also serve as a role model for involving civil society in effective responses to interference. Government measures alone will often have limited effect. Countering disinformation and election interference must draw on civil society representatives and initiatives as much as possible. Including grassroots organizations, academic institutions, and community leaders in efforts to combat disinformation can make them more effective, in part by helping bridge the gap between government actions and public trust, and by promoting greater transparency and accountability.

Taiwan’s efforts on this front have included exploiting the data curated by NGOs with information on years of election interference and information manipulation. Some organizations then use generative AI to experiment with debunking chatbots (backed by human verification), conduct collaborative fact-checking, and develop methods of linguistic and cross-platform analysis to identify disinformation more effectively and definitively.

Taiwan’s all-of-society approach actually entails fostering collaboration among many sectors of society. Fact-checking organizations, for example, share their expertise and skills with traditional media outlets. And NGOs hold workshops with teachers to promote media literacy skills in schools.

Cooperation With Social Media

Taiwan can learn from other democracies, too. It would benefit from looking abroad, especially to Brussels, for legal frameworks that allow cooperation with social media platforms since the EU may have the most advanced legislation to combat election interference by bolstering online transparency and accountability. Its Digital Services Act (DSA) and recently published guidelines for mitigating systemic risks for electoral processes impose special responsibilities on online platforms and online search engines with more than 45 million users in the EU. By holding platforms accountable for their content moderation, labelling practices, and enforcing greater transparency, the DSA aims to strengthen the resilience of democratic processes.

The act’s effectiveness will ultimately depend on the political will and capacity for enforcement, the ability to keep up with new developments in information manipulation, and the ability to foster collaboration among government, business, and society. Still, the DSA provides a useful framework and can serve as a reference for legislation in other democracies. It may also impact global regulation as multinational companies try to streamline compliance efforts across jurisdictions.

Tackling the Trust Gap 

Despite all the efforts, engendering trust in government authorities, democratic institutions, and reliable sources of information remains a challenge. Many measures to counter election interference, including those promoted by the DSA and related guidelines, require such trust to be effective. Without it, the proliferation of false and misleading information through social media platforms can create or reinforce information silos, polarization, and skepticism.

Radical transparency about the processes and mechanisms involved in addressing disinformation is needed for preserving or, where necessary, rebuilding trust. One way to achieve this is through sharing methodologies and information sources used in fact-checking, clarifying the criteria for determining the credibility of those sources, acknowledging potential mistakes or shortcomings, and making fact-checking more collaborative and inclusive. The Taiwanese NGO CoFacts explicitly encourages crowdsourcing of information verification and “fact-checking the fact checkers”. 

Collaboration Among Democracies

While distrust and polarization will remain challenges, most countries lack the resources to tackle them on their own. The burden will only grow as methods for election interference evolve and new threat actors emerge or expand. 

The pressure is on democracies to increase and accelerate processes for information sharing. Some of these processes, such as within the EU or the G7, already exist. These should be continually improved and expanded. Only by strengthening such mechanisms can democracies collectively strengthen defenses against the ever-growing threats to fundamental democratic principles.