Abstract

This article discusses the introduction of AI in the Taiwanese market, given its size and capacity, and stresses its recent regulatory measures implemented by the Taiwanese AI public policy in the resolution of domestic policy. I argue that this surge occurs amid the increasing international pressure of AI on the territory in light of the hegemonic stability that ties the United States (US) and the People's Republic of China (PRC) as regards the Taiwan Strait issue. 

Introduction

On February 3, 2025, Taiwan banned the use of the recent Chinese startup’s Large Language Models (LLMs), DeepSeek AI. Taiwan Premier Cho Jung-tai deemed that this AI should be banned in order “to ensure the country’s information security” (Reuters, 2025). On the one hand, one could say that this was a predictable reaction, given the tense relations between the PRC and the Republic of China (ROC). On the other hand, one could argue that it reveals Taiwan’s blind, indirect allegiance to the US or even a form of partial independence for the insular country, entangling domestic and international politics. This article introduces the contrasting entry of AI into the Taiwanese market, in light of its size and capabilities, and emphasizes the recent regulatory measures introduced by Taiwan's AI public policy to address domestic issues. I contend that this regulatory momentum is occurring amidst growing international pressure on AI within Taiwan, influenced by the hegemonic stability between the US and China concerning the Taiwan Strait issue.

The Maintenance of Leading-Edge Economic Assets

In the first place, I state that the baseline Taiwanese AI public policy has been to maintain its economic assets through the intelligent use of its current marketplace. With the aim of sustaining a comparative advantage in international trade and achieving economic growth, Taiwan has focused on semiconductor production since the 1970s (Tseng, 2024), alongside chips, sine qua non in the production of computing engines. This existing industrial environment has positioned Taiwan for a leading-edge role in AI in the modern digital economy—Taiwan outputs in mass one of the smallest chips, 3 nm, on the market (Reuters, 2023). Nonetheless, the maintenance of this oligopoly in the supply chain has required the integration of AI in the Taiwanese market through fiscal stimulus and regulation by the implementation of potent public policy by the Executive Yuan. 

In terms of fiscal stimulus, this plan considered in advance the size of the Taiwanese market: 95% of the companies in Taiwan are small and medium-sized enterprises (Reuters, 2023). This large sample of companies facilitated the integration of AI in Taiwan for the Executive Yuan through a competitive market and broader, easier collection of data for the implementation of legislative material. The Grand Strategy for a Small Country (2017-2021), settled upon under the Ministry of Science and Technology (MOST), provided for the investment of US$517.5 million to foster AI in the private sector (National Science and Technology Council, 2018), and the Ministry of Digital Affairs (MODA) created in 2022 the Enhanced AI Startup Investment Program, allocating US$320 million from the National Development Fund to stimulate R&D, taking advantage of the size of the Taiwanese market (Department of Digital Strategy, 2025). On the economic side, this encouraged companies to emerge as big players, as in the case of Appier, which raised US$80 million in 2019 (Goh, 2019), increasing the weight of the Taiwanese AI market. This innovation momentum was also enhanced through legal provisions: the AI Action Plan 1.0 (2018–2021) advocated sharpening and innovating in this market through deregulation and open access of firms to the market (National Science and Technology Council). Nonetheless, even though this plan appears to be free of regulation, it proves to be in opposition to the policies that will occur afterward; therefore, it is not only about upgrading the economic stimulus, but also a case of controlling domestic pressure. 

The Resolution of Domestic Pressure

In the second place, I argue that Taiwan’s AI public policy converged toward the resolution of domestic pressure through regulation of the market and governance over AI in the aftermath of the AI Action Plan 1.0. In fact, even though I mentioned it for economic stimulus, the MODA was also established to oversee the market and make a transition to the AI Action Plan 2.0 (2023–2026), which advocates for more market regulation and AI-based governance (Blablová, 2025). I contend that these regulations emerged amidst the ruling Democratic Progressive Party’s (DPP) intention to ensure the stability of domestic policy in the face of international pressure.

In that case, one needs to break down the false promise of Taiwan as an open-market AI innovator free from internal divide. There is indeed a gap in the perception of AI in the public discourse. While the DPP's public policy supports it, the opposition, mostly conservatives in the Kuo Min Tang (KMT), fear a rise in unemployment, social rigidity and totalitarianism (Wen et al., 2024). On December 27, 2024, opposition parties, holding a majority in the Legislative Yuan, even proposed a bill to cut spending on technology and redirect it to local municipalities (Lee, 2024), thereby hindering de facto the fiscal stimulus. Nonetheless, these political divisions appear negligible in the broader Taiwanese population: their perception mostly relies on their respect for science authority and knowledge (Wen et al., 2024). Overall, the population seems to advocate for more regulation over AI, while being more lenient if the latter touches upon national security policy toward China (Blablová, 2025). Therefore, through stronger regulation, the Taiwanese government managed to be in line with the public discourse’s incentives and preserve its market. The combination of these two dynamics may have ultimately reinforced trust in the DPP’s ability to sort out domestic issues and have a hermetic market, in order to address increasing trust in international AI pressure in light of the US-China AI competition. 

The US-China Hegemonic Stability in Light of AI

In the third place, I state that the incentives for the implementation of regulation in AI policies are the results of international pressure driven by the US-China competition over the Strait of Taiwan. Though AI is generally discussed in light of domestic politics, it is seldom envisioned in the theory of international relations, as its impact on the global order has yet to be estimated clearly. Nonetheless, some of the areas it impacts can be delineated, amongst others: AI reassesses the place of cybersecurity from the perspective of global inequalities and alliances (Pauwels, 2019)—the Global North develops more AI and can more easily pose a threat to the Global South; AI can enhance disaster and global health crisis management through the modelling of more precise scenarios based on machine learning (Meleouni and Efthymiou, 2023); AI can assist in the analysis of international law and support its proper implementation (Deeks, 2020). Even though the UN seeks to provide avenues for inter-state cooperation in these areas through the creation of a living UN Charter on AI (Pauwels, 2019), this remains a theoretical perspective. In fact, AI empowers the novel diplomatic interplay between non-state actors—tech industries—and state actors (Ndzendze and Marwala, 2023). In the case of Taiwan, this refers to the reliance of the Silicon Valley giants on Taiwan Semiconductor Manufacturing Company and its chips (Blablová, 2025). Therefore, each company becomes part of its host country’s projection capacities and assists the latter directly or indirectly. 

This leads to the discussion of the US-China rivalry and the issue of Taiwan in containing the hegemonic stability of both powers. The main contours of the hegemonic stability theory (HST) can be delineated as follows: it is “essentially the assertion that a single country can dominate a world order and create conditions for a more open and stable commercial environment than would be the case in a bipolar or multipolar order” (Ndzendze and Marwala, 2023). In that sense, Taiwan is the saddle point of the assertions of both powers: the fall of the ROC to the PRC would for sure assert the domination of the latter over the international order. The HST relies on three pillars that are being contested by the US and China, each of which implicates AI: the military projection, the economic power, and the battle of perception and knowledge (Ndzendze and Marwala, 2023). The military projection by both countries can be aided through access to Big Data and more accurate reconnaissance and intelligence devices. The economic power of AI can be ensured through investment in R&D in Taiwan’s AI industry. Through other lenses, AI could also be seen as an authoritative producer of knowledge. LLMs are at the vanguard of this shaping-knowledge process, due to their scope and their blind trust. As such, the development of DeepSeek as a direct competitor to ChatGPT 4.0, given their differences in datasets, may influence Taiwanese people's or American people’s perception about Taiwan (Dixon, 2025). The only lock that can maintain Taiwan as a sovereign state in this battle is the US’s perception. However, this perception depends on Taiwan’s perceived level of democracy (Chang, 2019), and there has been a drop in favorable perception of the US since  Donald Trump’s second term in 2025. One might think that Taiwan can develop its own LLMs to compete in the production of knowledge, yet the limited amount of traditional Chinese sources limits such ambitions. Thus, Taiwan has placed all its faith into the project Trustworthy AI Dialogue Engine (TAIDE) (Blablová, 2025), the “Taiwanese ChatGPT” in order to win the “semantic battle” and perhaps the war. 

Conclusion

The modern role of AI in Taiwan arises from its economic strengths and geopolitical situation. This development is a consequence of Taiwan's adoption of liberal stimulus policies, leveraging its large market and chip production. These policies were further merged into regulatory measures and the DPP's vote-seeking behavior in order to achieve domestic stability. As a result, Taiwan has attempted to face the international pressures from AI, amid US-China rivalry, being a hegemonic stability saddle point whose destiny depends on the White House’s fool and the engine.