When the Regulator Can't Meet
Taiwan has one of the world's most celebrated anti-disinformation ecosystems: a vibrant civil society of crowdsourced fact-checkers, a government that prefers memes to mandates, and a democratic culture hardened by decades of coordinated Chinese influence operations. Yet as the island prepares for its November 28, 2026 local elections — simultaneous votes across all 22 cities and counties that will fill some 11,000 elected positions — the structural weaknesses in its platform governance framework are becoming difficult to ignore.
The most immediate problem is administrative. In November 2025, Taiwan's Legislative Yuan rejected all four nominees to the National Communications Commission (NCC), leaving the country's primary communications regulator with only three sitting commissioners. Under its governing statute, the NCC requires four members for a quorum to hold formal sessions and issue binding decisions. Until the political deadlock between the minority DPP government and the KMT-TPP opposition is resolved, Taiwan's primary media and platform oversight body cannot make new regulatory policy of any kind.
The 2022 Attempt and Its Lessons
Taiwan has been in this position before. In June 2022, the NCC proposed the Digital Intermediary Services Act (DISA) — a comprehensive framework modeled closely on the European Union's Digital Services Act. It would have introduced transparency obligations, algorithmic accountability requirements, and content moderation standards for large online platforms operating in Taiwan, closing a regulatory gap that had left internet platforms almost entirely unregulated while traditional broadcasters faced extensive NCC oversight.
The steelman case for DISA was real: Taiwan faced documented foreign information operations, disinformation circulating during elections and public health emergencies, and platforms operating with no legal duty to disclose their algorithmic systems or enforcement practices.
The public reaction was swift and decisive. On the government's own online participation platform, more than 30,000 citizens registered formal opposition to the draft, versus fewer than 150 in support. Critics identified genuine structural flaws. The bill imposed identical compliance burdens on global platforms like Meta and on small domestic platforms like PTT and Bahamut, which operate on volunteer labor and could not afford the legal overhead. Its content-takedown triggers were vague enough to chill legitimate speech. And the sharpest objection: the law reached Taiwanese platforms but did nothing about disinformation originating from Chinese services like Weibo and Douyin, which sat entirely beyond Taiwan's jurisdiction — potentially disadvantaging domestic platforms while leaving the primary threat untouched.
By August 2022, Premier Su Tseng-chang suspended the bill. As of 2026, it has not been revived, and Taiwan has no dedicated platform regulation law.
An Anti-Fraud Law Doing Platform Work
The legislative gap has not gone entirely unfilled. Taiwan's Fraud Crime Hazard Prevention Act (FCHPA), enacted on July 31, 2024, creates binding obligations for online advertising platforms. Platforms must remove fraudulent advertisements within 24 hours of notification from authorities, verify advertiser identities using digital signatures or equivalent technologies, disclose whenever ads use deepfake or AI-generated content, and publish annual fraud-prevention transparency reports. The penalty structure is significant: standard violations draw fines of NT$500,000 to NT$10 million, rising to NT$100 million (approximately US$3.2 million) for severe breaches, with potential DNS and access restrictions for continued non-compliance.
Enforcement has begun. In May 2025, the Ministry of Digital Affairs issued Taiwan's first fine under the framework — NT$1 million against Meta for failing to disclose advertiser identities on Facebook. Deepfake celebrity endorsement scams dropped measurably on Facebook and YouTube following the law's entry into force.
But the FCHPA was designed to combat financial fraud, not political disinformation. It reaches fake investment advertisements impersonating public figures; it does not reach state-sponsored narratives designed to shape voter views on cross-strait relations. This distinction matters critically as November 2026 approaches.
The Xiaohongshu Ban as Governance Stress Test
The tension between democratic defense and proportionate regulation crystallized in December 2025 when Taiwan's Ministry of Interior ordered a one-year DNS block of Xiaohongshu — the Chinese lifestyle platform known internationally as RedNote — under Article 42 of the Anti-Fraud Act. The stated justification was financial: NT$114.77 million in fraud losses traced to the platform between January and November 2025, affecting more than 3 million Taiwanese users.
The ban exposed a fundamental governance mismatch. The legal instrument was the Anti-Fraud Act; the underlying policy concern was cognitive warfare. The result was a "Digital Homecoming" effect in which thousands of users turned to VPNs and alternative platforms, demonstrating that technical enforcement measures cannot resolve politically contested questions about cross-strait identity — and that repurposing fraud law for geopolitical objectives risks credibility in both domains.
The Threat Is Not Waiting
Taiwan's adversaries are not observing a pause. An OpenAI report published in February 2026 documented a PRC state-level AI cognitive warfare infrastructure with 300 or more operators per province employing over 100 documented tactics. Citizen Lab has identified at least 123 spoofed news websites mimicking legitimate Taiwanese outlets, mixing factual content with targeted falsehoods optimized for search engine distribution. Taiwan's Ministry of Justice, responding to the scale of the threat, designated AI disinformation as an independent election investigation priority for 2026 — the first time the category received separate treatment from general false-statement provisions.
What Taiwan Gets Right
Taiwan's civil society infrastructure is a genuine achievement and deserves recognition. Crowdsourced fact-checking organizations — CoFacts, the Taiwan FactCheck Center, and MyGoPen — provide rapid, distributed verification without centralized government removal orders. The government's "Humor Over Rumor" communications strategy deploys satire and accessible public messaging over legal mandates, treating disinformation as a public health problem requiring community immunity rather than a policing problem requiring state enforcement. Taiwan scored 79 out of 100 on Freedom House's Freedom on the Net 2025 assessment, classified as "Free" — a score that reflects genuine internet openness.
The Anti-Infiltration Act of 2019 provides a targeted legal instrument, criminalizing activities by individuals receiving support from hostile foreign forces for election interference, with penalties of up to five years' imprisonment. Used carefully, it addresses foreign-directed influence operations without requiring platforms to become disinformation adjudicators.
The Proportionality Argument
The 2022 DISA failure did not demonstrate that platform regulation is impossible in Taiwan — it demonstrated that the specific proposal was disproportionate. A workable framework would need to impose tiered obligations scaled by platform size and market power, reserve the heaviest transparency and algorithmic audit requirements for the largest global platforms, and acknowledge explicitly that cross-border Chinese platform operations fall outside its reach. The EU's own DSA, which DISA was meant to emulate, builds in exactly these proportionality mechanisms.
None of that legislative work can happen while the NCC lacks a quorum. Restoring the commission's capacity is not merely an administrative matter — it is a democratic infrastructure question. Taiwan's information environment in November 2026 will partly reflect whether its legislature can resolve a governance deadlock before the ballots are printed.