As Taiwan seeks to translate its semiconductor strength into broader leadership in AI, the question is no longer simply how to build better models, but how to govern the systems increasingly embedded across government, industry and society.In the latest episode of Taiwan Frontlines, produced by Taiwan Current News (TCN) in partnership with the German Marshall Fund of the United States (GMF), Isabel Hou (侯宜秀), the deputy minister at Taiwan’s Ministry of Digital Affairs (MODA), discusses the country’s emerging approach to AI governance, sovereign AI and the infrastructure she believes Taiwan needs for the AI era.Hou brings a legal and civic-technology background to the discussion. A National Taiwan University (NTU) law graduate, she worked as an attorney on technology and public-policy issues from 2006 to 2025 and served as secretary-general of the Taiwan AI Academy Foundation from 2021 to 2025.She was also involved in Taiwan’s g0v, a decentralized, grassroots civic-technology community, and the Legislative Yuan’s Open Parliament initiative.She became MODA deputy minister in September 2025, shortly before Taiwan’s Legislative Yuan passed the Artificial Intelligence Basic Act in December. The law was promulgated and took effect on Jan. 14, 2026.The AI Basic Act: a framework rather than a finished rulebookIn the latest Taiwan Frontlines episode, Hou described the legislation as the foundation for an ongoing governance process rather than a comprehensive rulebook that attempts to anticipate every future AI application.According to MODA, the act establishes seven core principles for government promotion of AI, including human autonomy, privacy protection and data governance, cybersecurity and safety, transparency and explainability, fairness and nondiscrimination, accountability, and sustainable development and well-being.Hou further explained that the framework operates on three levels. At the top are the seven broad principles; at the bottom, individual ministries and agencies determine the regulatory or policy responses appropriate to their respective domains.Between them is an “orchestration layer” involving AI risk taxonomy and assessment mechanisms. MODA Deputy Minister Hou discusses digital affairs and AI. (TCN) Hou added that MODA is now working with different agencies to identify AI use cases, assess potential risks and determine appropriate responses.For Hou, the process cannot be static. Because AI systems evolve rapidly, she stated that governance must be continuously adjusted through inter-ministerial coordination and engagement with practitioners.Sovereign AI: capability and controlThe conversation then turned to one of the most contested concepts in AI policy: sovereign AI.Hou rejected the idea that sovereignty requires Taiwan to independently manufacture every component of an AI system. Instead, she framed sovereignty around capability, control and accumulated knowledge.Her proposed scope includes data autonomy, controllable models, trustworthy governance, widespread application, and the ability to continuously learn and accumulate know-how.She described AI as a chain of control points: data at the input stage, model selection and computational infrastructure during processing, and the management and reuse of outputs.The objective, she said, is to create a “flywheel” in which knowledge generated through AI applications can become future inputs, allowing Taiwan’s AI ecosystem to learn and evolve.Three pillars for the AI eraHou also identified three pillars of the new AI infrastructure.The hardware pillar encompasses energy, chips and computing infrastructure. The software pillar covers data, models and evaluation mechanisms. The third pillar — equally essential in Hou's formulation — is governance: the framework through which Taiwan’s values are reflected in the development and deployment of AI.Together, she argued, these pillars extend Taiwan’s traditional technological strengths into a broader national AI capability. Hou called the three pillars a “new infrastructure for AI era” and said the concept was included in AI strategic guidelines released in Taiwan in July.For more in-depth coverage, tune in to Taiwan Frontlines on the NOWNEWS official YouTube channel.