Taiwan Counters FIMI – Governmental and Parliamentary Responses

Executive Summary

1. Taiwan’s national-level (governmental and parliamentary) responses to FIMI largely started in 2017 with government initiatives and preliminary research, marked by the first public debut of an Executive Yuan (central government)-level taskforce on countering disinformation in 2018, domain-specific legislations passed by the Legislative Yuan (national parliament) in 2019, COVID-19 responses in 2020, and a fundamental shift of governmental and parliamentary priority towards anti-fraud since 2022.
2. Taiwan’s government response to FIMI is decentralized both in terms of domain (domestic affairs, foreign affairs, national defense, economic development, agriculture, public health, etc.) and locality (central and local government and law enforcement agencies). National Institute of Cyber Security and Cognitive Warfare Research Center are two noteworthy central government institutions partially responsible for countering FIMI.
3. Taiwan’s legislative response to FIMI is also decentralized in terms of domain. Legal amendments passed in 2019 largely unified domain-specific provisions in existing laws, penalizing the dissemination of damaging rumors or falsehoods in specific domains such as disaster prevention and communicable disease control.
4. Legal amendments to curb foreign infiltration are key to countering FIMI in Taiwan. Amendments to the Criminal Code (2019) rectified the scope of treason to include adversaries in (mainland) China, Hong Kong, and Macau. Anti-infiltration Act (2019) is the sole new legislation passed related to countering FIMI, banning foreign interference in Taiwan’s democratic processes.
5. Taiwan currently lacks viable pathways towards legal regulations on digital services in terms of countering FIMI. Two major legislative proposals on digital services regulation have failed to enter deliberation at the Legislative Yuan in 2018 and 2022.
6. Taiwan’s law enforcement has filed increased numbers of charges in 2019 and 2020 on individuals spreading rumors in violation of the Social Order Maintenance Act. Concerns have been raised by human rights organizations and activists in Taiwan.
7. Taiwan has multiple societal and systemic vulnerabilities and remains vulnerable to FIMI on all major narrative domains related to statehood, including domestic economy, government performance and legitimacy, national defense capabilities, international relations and trade, history and culture.
8. Lack of pre-legislative multistakeholder deliberation, post-legislative scrutiny, holistic assessment on political and non-political effect of FIMI, and framework to increase healthiness of the information environment are the main shortcomings of national and international responses to FIMI.
9. Taiwan-United States relations, legitimacy of Taiwan’s elected government, and major legislative proposals are likely topics for upcoming FIMI activities in Taiwan.
10. AI-powered content production, astroturfing, and other types of human mimicry and automation could transform FIMI operations in terms of quantity, diversity, and manufactured authenticity.

Recommendations to Taiwan’s Government and Parliament

1. Empower expert-driven research, technological development, procurement, and program development at National Institute of Cyber Security, Cognitive Warfare Research Center, and other government institutions to increase effectiveness in countering FIMI.
2. Respect civil society independence, avoid control and direct coordination. Build support systems, including independent funding structures, for diversity and sustainability of fact-checking, information environment research, data journalism, digital civics education, scientific communications, and other non-government activities related to countering FIMI, facilitating public discourse, and improving information environment healthiness and robustness.
3. Build trustworthy partnerships and networks on transparent mechanisms for information sharing and verification among government and non-government actors.
4. Compile best practices and develop transparent evaluation processes and relevant competency to validate government and non-government FIMI research according to best practices and processes.
5. Establish public opinion baseline and assess effectiveness of FIMI via regular public opinion surveys conducted by reputable polling organizations and shared publicly in both survey data and methodology.
6. Develop analytical and risk assessment framework for FIMI narratives and behaviors, as well as other threat activities to inform appropriate and proportional response. Adapt existing definitions and frameworks to local context, limitations, and strengths.
7. Clarify rumors, provide facts, offer evidence-based assessment, inform public discourse. Never conflate official clarification with fact-checking. Strengthen public literacy on key policy domains vulnerable to FIMI – public health, food safety, supply of electricity and everyday goods, election integrity, semiconductor, biotech, arms procurement, national defense, and foreign affairs.
8. Refrain from unsubstantiated or unsound accusation of FIMI, cognitive warfare activities, or collusion with foreign adversaries in governmental, parliamentary, and partisan discourse. Separate FIMI and cognitive warfare from domestic party competition.
9. Address legitimate criticism and skepticism, improve government performance and accountability, increase transparency and meaningful participation in policymaking and lawmaking, introduce policies and programs in reducing socioeconomic, ethnic, gender inequalities, partisan divisions, and other vulnerabilities linked to the effectiveness of manipulative narratives.
10. Build strategic communications capabilities into key governmental departments. Co-develop with civil society evidence-based public communications guidelines and implement training programs for government officials, civil servants, and parliamentary staff, and party staff. Co-host public communication workshops at the local level among government offices, party offices (affiliated with national or local lawmakers), and local communities.
11. Co-host non-partisan, inclusive, participatory, and transparent national forums at the Legislative Yuan with governmental, parliamentary, academic, media, educational, civil society experts, civic tech and civic hackers, and other non-governmental actors for multistakeholder engagement, deliberation, and consensus-building on sensitive legislations such as foreign agent registration and digital services regulations, as well as other legislations and policies related to countering FIMI, facilitating public discourse, and improving information environment healthiness and robustness.
12. Clarify the definition of rumor in the Social Order Maintenance Act.
13. Legislate for proactive government, platform, and media corporate transparency in content moderation, account takedown, personal and financial information inquiries, and other regulatory measures related to countering FIMI.
14. Design and institutionalize post-legislative scrutiny as an independent function of the Legislative Yuan.
15. Actively create space for Taiwan’s international participation. Actively participate and contribute Taiwan’s collective expertise and perspective in international forums on digital platform accountability and transparency, internet governance, and AI governance through governmental and parliamentary diplomacy in collaboration with civil society.

Introduction

Taiwan’s governmental and parliamentary responses to FIMI largely started in 2017, signified by the addition of controversial information clarification (爭議訊息澄清) sections on government websites including one in top government portal implemented by the National Development Council (NDC), Taiwan’s top policy-planning agency, and multi-stakeholder convenings among platforms, academia, and civil society hosted by the National Communications Commission (NCC), Taiwan’s independent body governing telecommunications. At the time, Taiwan’s Executive Yuan (EY) instructed ministerial responses while expressing no intention in initiating legal amendments for content regulation. Neither of NCC’s drafts to amend the existing Telecommunications Management Act (電信管理法) and legislating a new Digital Telecommunications Act (數位通訊傳播法) explicitly mentioned any FIMI-related terms. However, analysis on “fake news” (假新聞) by the Organic Laws and Statutes Bureau (法制局) of Legislative Yuan (LY) was an early sign of future legislative measures.

Topic of FIMI first entered Taiwan’s legal framework in late 2018, around Taiwan’s local elections (24th November 2018). Both the LY and EY have conducted internal research and published relevant information in 2018 on national-level activities countering disinformation (假訊息). An EY-level special taskforce (專案小組), led by then-Minister without Portfolio Luo Ping-Cheng (羅秉成), established an initial 3-element framework for evaluation: malicious intent, falsified content, and harmful results (惡, 假, 害), and a 4-stage framework for response: detect, debunk, contain, and discipline (識假, 破假, 抑假, 懲假).

State Actors Countering FIMI

Government agencies, depending on their legal designation, have a variety of roles countering FIMI: research, policymaking, institution-building, lawmaking, as well as direct implementation of FIMI countermeasures.

Multiple central and local government agencies in Taiwan over recent years have procured public opinion monitoring services from commercial entities, likely for manual or keyword-based detection of falsehoods and rumors related to the domain or locality of the procuring agency.

Clarification of falsehoods or rumors, depending on the subject matter, falls to a wide range of government agencies in the central government including the Legislative, Executive (including the military), and Judicial Yuans (JY), Ministries, Councils, Administrations, Bureaus under EY, as well as city and county governments and its various bureaus.

At the presidential level, Taiwan’s top intelligence agency the National Security Bureau (NSB) has indicated its use of databases and analytical tools to detect PRC threat activities in realtime. At the EY (central government) level, a special taskforce led by then-Minister Luo Ping-Cheng made its first public debut in December 2018, positioned itself to be responsible for cross-departmental coordination within the government to counter disinformation. Around this time, national police and military have both set up dedicated units for countering disinformation. Criminal Investigation Bureau (CIB) established Disinformation Investigation Unit (查緝假消息專責小組) in October 2018 during local elections, later elevated as Disinformation Investigation and Response Unit (假訊息查處專責小組) under National Policy Agency (NPA) in January 2019. Ministry of National Defense (MND) announced in May 2019 the establishment of a military Disinformation Rapid Response Unit (反制假訊息快速處理小組) to strengthen public trust on the military, and fortify psychological defense of military personnels.

National Institute of Cyber Security, formally established in January 2023, is responsible for national-level cyber threat analysis and response, including technical research and development on countering information manipulation and anti-fraud. Cognitive Warfare Research Center, formally established in January 2024, is set to combine cross-strait situational analysis, domestic security, and cybersecurity capabilities within the Ministry of Justice Investigation Bureau (MJIB) to strengthen law-enforcement response to FIMI.

Parliament’s Role Countering FIMI

Taiwan’s national legislative body, the Legislative Yuan (LY) mainly has three roles countering FIMI: deliberation of government-proposed bills related to countering FIMI, proposition and passage of bills countering FIMI, and formal inquiries on government activities countering FIMI.

First, Taiwan’s LY deliberates, modifies, and passes multiple EY-proposed amendments to existing laws related to countering FIMI. Deliberation of amendments take place in any of eight standing committees depending on relevance to the subject matter.

In December 2018, EY finalized and proposed its first round of amendments to LY to seven existing laws related to countering FIMI, including Disaster Prevention and Protection Act (災害防救法), Food Administration Act (糧食管理法), Agricultural Product Market Transaction Act (農產品市場交易法), Communicable Disease Control Act (傳染病防治法), Food Safety and Sanitation Act (食品安全衛生管理法), Nuclear Emergency Response Act (核子事故緊急應變法), and Radio and Television Act (廣播電視法). LY passed five of the seven proposals in May and June of 2019.

Secondly, members of LY and party caucuses propose, deliberate, modify, pass amendments to existing laws, and legislate new laws related to countering FIMI.

In May 2018, 19 LY members led by Chiu Chih-Wei (邱志偉) and Su Chen-Ching (蘇震清) proposed amendments to the Social Order Maintenance Act to expand the scope of existing penalty on deliberate circulation of rumors (謠言) to include fake news and falsehoods (假新聞, 假消息). The proposal received wide criticism and failed to pass the second reading. In May 2019, LY passed amendments to Taiwan’s Criminal Code, originally proposed by Wang Ting-Yu (王定宇) and 17 legislators from the ruling Democratic Progressive Party (DPP), to rectify the definition of treason to include collusion with personnel of the PRC, including Hong Kong and Macau. In November 2019, DPP caucus leader Ker Chien-Ming (柯建銘) and two DPP legislators proposed Anti-infiltration Act (反滲透法). LY passed the proposal on 31st of December, 2019, and then-President Tsai Ing-Wen signed it into law on 15th of January, 2020.

A key factor for successful legislative measures countering FIMI is the alignment of the Presidential Office, EY, and LY majority, all three held by the ruling DPP between 2016 and 2024. Results of the 2024 election (in January) have absolved such alignment between the executive and the legislative branches of the central government of the next four years (2024-2028).

Thirdly, members of LY practice their constitutional power of formal inquiries (質詢) during committee proceedings to request information and demand further actions from relevant government agencies. Government officials are legally required to be physically present and respond to legislative inquiries. Committee proceedings are legally required to be live-streamed via dedicated TV and YouTube channels for accountability and public discourse.

Timeline

Dynamics and Evolution

Vulnerable Narrative Domains

IORG uses open-source and proprietary natural language processing (NLP) and machine-learning (ML) technologies for identifying FIMI narratives and behaviors in the Mandarin-language information environments across platforms and localities including Taiwan and other Mandarin-speaking communities. Narratives tracked and documented by IORG are classified according to the targeted country, domain, and their affective triggers.

Taiwan, the United States (US), Japan, and the People’s Republic of China (PRC) are amongst the most targeted countries by Mandarin-language FIMI narratives. Such narratives collectively reinforce defeatism towards Taiwan, skepticism towards the US, animosity towards Japan, while strengthening the dual perception of power and benevolence of the PRC.

Domestic economy, government performance and legitimacy, national defense capabilities, international relations and trade, history and culture are domains commonly exploited by FIMI narratives shaping the image of Taiwan as a failing nation in every aspect.

Inequality is a common affective trigger of FIMI narratives circulating in Taiwan. Existence of inequalities and perceived injustices between socioeconomic strata, traditional male-female roles, and disparity of power and status among ethnic-lingual groups, as well as affirmative measures intended to mitigate such gaps can both be exploited to magnify social divisions and exacerbate mistrust.

Individual or collective self-perception of smallness, weakness, and lack of agency for self-determination is another common affective trigger for FIMI narratives in Taiwan. Probable causes of this pessimism include the legacy of a China-centric education system, US recognition of the PRC in 1979, long-term wrongful exclusion of Taiwan from international affairs and security reassurances, as well as individual and collective economic anxiety.

Systemic Vulnerabilities

Taiwan’s society is dominantly Han-immigrant descendants and its political and cultural discourse dominated by Mandarin-language content, largely a result of early Han immigration, post-war refugee, and oppressive policies imposed by the Chinese Nationalist (KMT) military dictatorship before democratization began around 1990.

Mutual intelligibility between Taiwanese Mandarin and the Standardized Mandarin used in the PRC (Putonghua) lowers the barrier of entry for tailoring content production to the preferences of the Taiwanese audiences. Mandarin expressions used in the PRC, sometimes lacking corresponding ones in Taiwan, such as those in popular culture or professional disciplines, can be perceived as novelty or necessity in Mandarin-based communications, drawing Taiwanese audiences closer to PRC’s linguistic and value systems.

Taiwan’s reduced yet still significant economic dependency on and interconnectedness with the PRC and the common conception that Taiwanese businesses and Taiwan as a whole must depend on China as the only (easiest) option for economic gain are two mutually-reinforcing concepts in Taiwan’s collective consciousness. Polls have indicated an incline of public awareness of the PRC as the main security threat to Taiwan while a declining yet still significant portion of Taiwan’s public maintain their support for cross-strait economic integration.

Taiwan’s legacy of the former KMT military dictatorship and long-standing polarizing partisanship provide ready access to fissures between party and national identities, and abundant opportunities for manipulative actors to mobilize emotion, consolidate support, and prioritize party interest over national sovereignty.

Profit-driven sensationalism among commercial mass media, weak public media, low quality standard for news, as well as normalized partisan perception of news inhibit the media sector from contributing to reasonable public discourses and a healthy information environment.

Shortcomings of National and International Responses

Nationally, the lack of transparent post-legislative scrutiny on legislations related to countering FIMI have prevented public review and evidence-based public debate on the efficacy of such legislation.

Nationally, the lack of sustainable public consultation and multistakeholder deliberation process for sensitive legislation including foreign agent registration and digital services regulation have prevented forming of societal consensus around such important topics and have allowed the absence of important legislation countering FIMI.

Nationally, uncritical emphasis of the vibrancy of civil society response to FIMI have allowed inaction of government officials and agencies responsible for countering FIMI. Internationally, focus of regional and national government responses to FIMI have limited the scope and available methods of those responses.

Nationally and internationally, the lack of awareness and evidence-based framework on mapping both political and non-political negative effects of information manipulation have limited the diversity of responses needed towards a multidisciplinary, multistakeholder holistic solution. Non-political effects of information manipulation include negative effects on youth development, mental health, privacy, cognitive autonomy, and social trust.

Nationally and internationally, the lack of awareness and evidence-based framework on identifying attributes of reasonable public discourses and healthy information environments have framed activities as reactive remedial measures rather than proactive constructive measures.

What’s Next

Upcoming Topics

Results of the US elections in November 2024, major US policy proposals, any Taiwan-US joint security measures, as well as any event that could be interpreted as a US withdrawal from regional and global security are highly likely to be high-priority topics of US Skepticism narratives, categorized by IORG’s milestone report in 2023.

Anti-infiltration, anti-corruption, constitutional court appointments, legislative proposals, as well as other legitimate executive measures could be framed as “Green Terror” (綠色恐怖), falsely equating the current DPP government to the dictatorial KMT government responsible for the forty-year-long White Terror (白色恐怖). (Green is the representative color of the current ruling party DPP.)

PRC military drills around Taiwan and PRC activities in the West Pacific internationally, as well as legislative proposals by EY or LY and prospective referendums proposed or backed by political parties domestically, if any, are likely topics for new FIMI narratives in Taiwan before the next election campaigns towards the end of 2026.

Threat to Support for Ukraine

There is no evidence to suggest a reduced level of support for Ukraine and Ukraine’s victory among Taiwan’s public.

An absence of Ukraine and the Russian invasion against Ukraine in mainstream media coverage and public discourse could result in reduced sense of urgency and increased apathy among Taiwan’s public towards Ukraine’s victory.

Narratives framing US and international support for Taiwan and Ukraine as mutually exclusive, such as ones observed by IORG, could raise skepticism among Taiwan’s public towards the US and Ukraine and reduce public support for Ukraine.

New Informational Threats

Continued instrumentalization of Taiwanese and other non-PRC individuals as vectors of FIMI entangles foreign interference with legitimate democratic discourse and poses significant challenges to FIMI countermeasures. The mix of domestication and organic domestic occurrence of FIMI narratives remain the most significant threat to Taiwan’s democratic resilience.

Manipulative narratives based on alleged leaked audio recordings, alleged public opinion polling results, or other fabricated or unverifiable factual elements diminish the significance of evidence, normalize haste conclusions with insufficient factual support, “de-normalize” expectation for reason and civility, and degrade the quality of public discourse.

Declining levels of credibility, authenticity, and connectedness in the information environment result in declined public trust and participation, which reduces feasibility of reasonable public discourse online, and legitimacy of (digital) democracy at-scale.

AI’s Role

New and accessible generative AI technologies have enabled cheaper, faster, multilingual, multimodal, at-scale mimicry of human content creation, diverse persona, and authentic interaction. Programming co-pilots and AI-generation of computer programs has significantly lowered the barrier of entry to software engineering and automation.

Integration of AI-enabled capability of human mimicry and automation could transform the operations of existing text-based content farms, video account clusters, astroturfing, social engineering, identity theft, and other FIMI and cyber threat activities to further confuse, manipulate, misrepresent, flood, and jam citizen participation in public discourse, as well as threaten their digital security.

Multiple cases were identified by Taiwanese and international journalists and researchers where AI technologies were used for election interference. (TFC 2023) (TFC 2023) (Wang 2024) (Microsoft 2024) For Taiwan’s 2024 elections, unverified audio recordings and “cheap-fake” videos (edited or altered with less sophistication) were more impactful than deepfake videos.

Researchers and technologists have long used AI to enable machine understanding of human language and behavior. Large language models (LLM) have the potential to enhance this process, enabling more precise and holistic understanding of public discourse and the ecology of information environments. Continued technical advancements by industry, academic, and civil society can make AI work for the betterment of the information environments and democratic societies.

Conclusions

Taiwan’s national-level activities countering FIMI largely started in 2017, the second year of President Tsai Ing-Wen’s first term, with government initiatives and preliminary research. At the end of 2018, an EY-level taskforce on countering disinformation made its first public debut, paving the way for legal amendments passed by the LY in 2019 to expand legal regulation on rumors and falsehoods regarding key areas of public affairs such as public health and food safety. At the same time, legal amendments were also passed to regulate cross-strait activities related to national security. The Anti-infiltration Act passed on the last day of 2019 concluded the year of legislation against FIMI in Taiwan.

The COVID-19 pandemic from 2020 occupied much of the first half of President Tsai’s second term. Daily CECC press briefings at 2pm became the source of reliable information for much of Taiwan’s public. Government use of humor and memes (especially Zong-chai the shiba-inu) in its strategic communications gained public acceptance and international acclaim while vaccine-related rumors and falsehoods were widespread. If the establishment of government agencies and institutions showed a national-level recognition of countering FIMI as integral to Taiwan’s national cyber and physical security, the fundamental shift towards anti-fraud since 2022 reflected a shifting national priority towards a more urgent threat, and the limitation to what the government and parliament could have accomplished.

Rumors, falsehoods, conspiracies, and unsubstantiated opinions have and always will be a part of liberal democracy. This long-existing vulnerability of Taiwan’s open society has been exploited by the PRC as one of its impactful means towards its conspicuous goal to annex Taiwan and demolish Taiwan’s hard-fought democratic way of life. In Taiwan, partisan polarization, social division, and a much compromised information environment have allowed malign actors to gain from discord, mistrust, and hate. Amongst many challenges, the entanglement of FIMI and legitimate domestic discourses remains the most significant threat to Taiwan’s democratic resilience.

When it comes to the betterment of Taiwan’s democracy, “no one is an outsider” (沒有人是局外人). Competent government effort, independent and credible civil society, as well as accountable public-private partnerships (PPP) are three keys to safeguarding basic civil liberties, strengthening digital civic literacy (in information, tech, privacy, data, cybersecurity, AI, civics, and public participation) among Taiwan’s public, civil servants, educators, and profession sectors, revitalizing Taiwan’s media sector, and increasing Taiwan’s national resilience to FIMI.