The Hoover Institution’s Project on Taiwan in the Indo-Pacific Region hosted “The FIMI Challenge: Countering Foreign Information Manipulation and Interference in Taiwan and the United States” on March 9, 2026, at Stanford University’s Herbert Hoover Memorial Building. You-Hao Lai, Deputy Director of DSET’s Democratic Governance Program, joined experts from Taiwanese civil society and the US academic community to examine the governance challenges and policy responses to foreign information manipulation in both Taiwan and the United States.
Taiwan’s Platform Governance Challenge and a Way Forward
In his presentation, Lai used Taiwan’s recent ban on Xiaohongshu (Rednote) as a case study to examine the structural challenges facing the Taiwanese government in addressing risks posed by Chinese platforms. He noted that the criticism prompted by the government’s reliance on the Fraud Crime Prevention Act to ban Xiaohongshu reflects a threefold governance dilemma: existing criminal penalties are ineffective against state-level FIMI; partisan divisions along national identity lines and the legacy of authoritarian rule make platform regulation politically difficult to advance; and Taiwan’s limited market size prevents it from imposing EU-style comprehensive regulatory obligations on foreign platforms.
On the way forward, Lai proposed two shifts in governance thinking. First, drawing on the operational safeguards in the US TikTok Deal, he argued that Taiwan’s regulatory approach should shift from ownership to operation—for example, using requirements such as data localization, algorithm oversight, and independent third-party audits as conditions for market access. Second, he suggested drawing insights from the EU’s FIMI framework, arguing that Taiwan should consider redirecting its regulatory focus from the truthfulness of content to the identification of manipulative behavioral patterns, in order to avoid triggering the political backlash associated with content censorship.
Lai also shared findings from his recent report, “The Authoritarian Gaze: China’s Global Data Reach and the Systemic Risks to Democracy,” identifying consistent data transfer pathways to China embedded in the privacy policies of ten Chinese AI services. Combined with China’s authoritarian data governance framework—which grants the government broad, largely unchecked authority to access data—these practices pose systemic risks to democratic societies. He emphasized that effectively governing Chinese platforms requires robust personal data protection, cybersecurity, and cross-border data transfer regulations, grounded in rigorous, evidence-based threat assessment rather than relying on ill-suited existing laws as stopgap measures.
Firsthand Observations from Taiwan’s Civil Society
In the first session on “Civil Society, Private Firms and FIMI,” Jerry Yu, Senior Analyst at Doublethink Lab, shared findings from the organization’s observation project on the 2024 Taiwanese presidential election, focusing on cases of FIMI. The research found that on TikTok, some influencers who typically post beauty and fashion content used identical materials and scripts to disseminate politically negative narratives, such as claims about the Taiwanese government bringing in Indian migrant workers and allegations of election fraud. AI-generated videos promoting conspiracy theories about former President Tsai Ing-wen were also widely circulated across major social media platforms. Survey results showed that frequent TikTok users were more likely than non-users to agree with narratives expressing distrust toward the government. Comparing content differences between TikTok and the Chinese version of Douyin will be an important direction for future research.
Wei-Ping Li, Research Director at FactLink and researcher at the University of Maryland, noted that Taiwanese society has successfully warded off much disinformation during key political moments over the past years, including the rapid containment of election integrity rumors on the night of the 2024 presidential election. However, with advancements in AI and changing socio-political conditions, Taiwan now faces a tougher battle: a more sophisticated propaganda campaign in which large amounts of AI-generated content aim to deepen societal divisions and erode public confidence—a shift from disinformation to propaganda warfare. Li called for a comprehensive approach, including monitoring propaganda patterns, understanding their effects on audiences, urging platforms to be transparent with their data and disclose more about content creators, and equipping users with more tools to identify problematic messages.
US-China Tech Competition and the Global Spread of Open-Source AI Models
Graham Webster, Research Scholar at Stanford University’s Center for International Security and Cooperation and Editor-in-Chief of the DigiChina Project, presented research on the global proliferation of Chinese open-source AI models. He noted that Chinese open-source models have matched or surpassed their US counterparts in both performance benchmarks and global download volume, and urged policymakers to distinguish whether risks stem from AI capabilities themselves, the openness of the models, or their connection to the Chinese system—and to adopt application-specific risk mitigation measures accordingly.
The symposium was organized by Kharis Templeman, Research Fellow at the Hoover Institution and manager of the Project on Taiwan in the Indo-Pacific Region. Larry Diamond, a leading scholar of democracy, chaired the first session. The forum facilitated exchange between Taiwan and the United States on platform governance, information environment resilience, and civil society responses, with in-depth discussions on the regulatory challenges posed by platforms from authoritarian states, the tension between content regulation and freedom of expression, and the emerging challenges of AI-driven information manipulation.
