The Taiwan Ministry of Digital Affairs (moda) is a cabinet-level ministry responsible for digital policy, infrastructure, regulation, and deployment, established on August 27, 2022. Its founding minister was Audrey Tang (2022–2024), succeeded by Huang Yen-nun (2024–present). The ministry's AI-related work centers on participatory policy deliberation, information-integrity efforts tied to Taiwan's exposure to disinformation, and support for the domestic open-source language model TAIDE.
| Established | August 27, 2022 | |
| Founding Minister | [[audrey-tang | Audrey Tang]] (2022–2024) |
| Current Minister | Huang Yen-nun (2024–present) | |
| Home | moda.gov.tw |
Mandate
moda consolidated previously fragmented digital authority into a single cabinet-level ministry. Its remit spans digital infrastructure (broadband, rural connectivity, and disaster-resilience networks); cybersecurity, consolidated under moda's Administration for Cyber Security; telecommunications regulation, which merged authority that had been split across earlier bodies; digital-government services such as one-stop digital services, identity, and forms; civic technology, including participatory-policy infrastructure; and AI policy and deployment, which encompasses support for TAIDE, the Alignment Assemblies, and information-integrity work.
AI uses and deployments
Participatory deliberation with Pol.is
moda uses Pol.is, a consensus-surfacing wiki-survey platform, for public consultation on regulatory decisions. On some topics the platform has engaged as much as half of Taiwan's population. Eric Schmidt's Democracy 2.0 essay cites the deployment as a proof-of-concept for scaling deliberative democracy.
Alignment Assemblies
Between 2023 and 2024, moda collaborated with the Collective Intelligence Project on a series of Alignment Assemblies. These included Ideathon events in Taipei and Tainan, structured as deliberative workshops on AI-era priorities; Recursive Public (November 2023), a participatory AI-governance deliberation; and the Information Integrity Assembly (March 2024), in which hundreds of thousands of randomly invited citizens deliberated on platform AI labeling, provenance, and AIEC assessment criteria ahead of Taiwan's 2024 presidential election. See Alignment Assemblies and Collective Constitutional AI.
TAIDE
moda supports Taiwan's Trustworthy AI Dialogue Engine (TAIDE), an open-source large language model built by Taiwan's National Applied Research Laboratories. TAIDE uses a constitution derived from the 2023 Taiwan Alignment Assembly for alignment-tuning. The ministry frames the project as part of Taiwan's sovereign-AI strategy, positioning a domestic model as insurance against dependence on US or Chinese frontier models. See AI Sovereignty.
AI Evaluation Center (AIEC)
The AI Evaluation Center is Taiwan's emerging AI-evaluation infrastructure, modeled in part after the UK and US AI Safety Institutes (UK AI Safety Institute (AI Security Institute), US AI Safety Institute (NIST AISI)). moda coordinates its evaluation criteria, including the information-integrity criteria that emerged from the 2024 Alignment Assembly.
Response to AI-assisted intrusion
On August 13, 2026 moda said it had detected an "abnormal attack" on Taiwanese government agencies during July, with its National Institute of Cyber Security issuing warning alerts from July 20. The ministry said investigators traced an overseas source and identified a hybrid method combining manual operations with AI agent-assisted attacks, naming Open Claw among the tools involved. The statement did not name China (Source: reuters.com).
The ministry's statement followed an account published the previous day by the Israeli security firm Dream, which reported that a team of AI agents had extracted scores of officials' passwords, taken personnel records from Taiwan's justice ministry, and scanned its nuclear safety agency for vulnerabilities over a four-day period (Source: reuters.com). See Autonomous cyber-agents.
Strategic positioning
moda's public-facing work emphasizes AI for democracy rather than AI regulation, drawing on Taiwan's status as a small state with an established civic-technology tradition. Its work is presented as distinct from the US, EU, and Chinese regulatory models, in contrast to the top-down approach reflected in China — Interim Measures for the Management of Generative AI Services and China — Provisions on the Administration of Deep Synthesis Internet Information Services. Information integrity is a central concern, driven by Taiwan's exposure to Chinese disinformation operations, and the open-source TAIDE model is positioned as a hedge against reliance on external frontier models. moda is frequently grouped with Asia-Pacific regulatory peers such as Singapore's Singapore MGF GenAI and Infocomm Media Development Authority (IMDA, Singapore).
The Tang/Siddarth/Huang essay in Digitalist Papers Vol. 1 covers moda's Alignment Assembly work at length.
Relationships
- led-by: Audrey Tang (founding minister).
- collaborates-with: Collective Intelligence Project (CIP) / Alignment Assemblies and Collective Constitutional AI.
- regulates: Taiwan telecoms, digital services, cybersecurity.
- deploys: Pol.is, TAIDE-derived systems, participatory digital infrastructure.
- related: UK AI Safety Institute (AI Security Institute) / US AI Safety Institute (NIST AISI) — AISI peer network.
- related: AI in Elections and Democratic Institutions / Liar's Dividend / Synthetic Media / Deepfakes — information-integrity work.
- related: AI Sovereignty — TAIDE fits this concept.
- related: Singapore Model AI Governance Framework for Generative AI (2024) / Infocomm Media Development Authority (IMDA, Singapore) — Asia-Pacific regulatory peers.