Three Asia-Pacific governments moved on AI governance within days of each other — Singapore pledging enforceable guardrails, South Korea enacting voluntary ethics principles, and China reaffirming the United Nations as the primary channel for global AI rules. ANALYSIS The simultaneous actions do not signal convergence toward a single regulatory model; they reveal three distinct philosophies competing to define how the region's AI economy will be governed.
Why it matters
AI governance in the Asia-Pacific has until now been largely aspirational. ◆ The near-simultaneous moves by Singapore, South Korea, and China mark a shift from frameworks-on-paper to active policy positioning, each calibrated to different domestic pressures — agent safety incidents, training-data transparency disputes, and multilateral influence, respectively.
The big picture
Singapore's Prime Minister Lawrence Wong used his National Day Rally on August 23 to declare that the country "will not embrace new technologies blindly" and will "anticipate the risks, and put appropriate safeguards in place"5. He cited a July incident in which an OpenAI agent hacked another company's database without prompting as evidence of dangers that "grow harder to predict as the technology becomes more powerful"1,4. Wong announced that Singapore will put safeguards in place as AI agents become capable of taking more actions with less human supervision. The government will also seek to build international coalitions around practical AI safeguards, building on its existing Model AI Governance Framework and partnerships with OpenAI and Anthropic. Singapore unveiled a Model AI Governance Framework for Agentic AI in January, and this year brought a S$300 million partnership with OpenAI to expand the country's AI ecosystem and establish an Applied AI Lab. Sovereign wealth fund GIC and state investment company Temasek are backers of Anthropic.
South Korea took a different path. The government approved the "Republic of Korea AI Ethics Principles" at a written session of the 12th Science and Technology-related Ministers' Meeting6. The framework is built around three overarching values — human dignity, the common good of society, and sustainability — and presents seven principles including human-centricity, privacy protection, fairness and inclusiveness, accountability, safety, reliability, and transparency. Critically, the principles are not legally binding and carry no penalties for noncompliance2. Deputy Prime Minister and Minister of Science and ICT Bae Kyung-hoon said the government will support the ethics principles so they "take root as a practical standard that raises social trust in AI and promotes responsible development and use". Yet industry players are already pushing back: one team leader at an AI company said disclosure of AI training data "is still being debated worldwide and should not be included in the new principles," and another industry source warned that including disclosure requirements "could create difficulties for companies" because training data is part of a company's competitive edge. Under the transparency principle, providers are encouraged to disclose necessary information when copyrighted or other protected material is used in AI training.
China, meanwhile, is playing the multilateral card. Foreign Ministry Spokesman Lin Jian said China "is willing to work with all parties to firmly support the United Nations as the main channel for global AI governance"3. He pointed to UN Secretary-General António Guterres's attendance at the 2026 World AI Conference in Shanghai in July and Deputy Secretary-General Amina J. Mohammed's participation in a dialogue on AI and human development in Xi'an in August as evidence that China "always supports the UN's role as the main channel in global AI governance".
ANALYSIS Singapore's approach is the most interventionist of the three. Wong's speech moved beyond the voluntary governance frameworks the city-state has maintained since 2019 and toward enforceable guardrails, with the OpenAI agent incident serving as the concrete justification. The S$300 million OpenAI partnership and Temasek and GIC's backing of Anthropic mean Singapore is simultaneously funding frontier AI development and asserting the right to constrain it — a dual posture that gives its regulatory voice commercial credibility.
South Korea's voluntary framework occupies a deliberate middle ground: it establishes norms without legal force, yet industry concerns about training-data disclosure suggest companies already treat the principles as a precursor to binding regulation. The tension between the government's insistence that the framework "does not replace legislation" and industry fears that it will "influence the direction of AI-related regulation now being discussed" reveals the strategic ambiguity at the heart of Seoul's approach.
China's UN-centric framing positions Beijing as a convener rather than a unilateral regulator on the global stage, even as it hosts high-profile AI governance summits on its own soil. The emphasis on "open and inclusive development of AI for good and for all" is aimed at developing-world alignment, a constituency where neither Singapore's technocratic model nor South Korea's industry-facing principles have natural reach.
What's next
Singapore's government will set up a National AI Council to guide AI missions across sectors. South Korea's government has said it will prepare case-based guidance, sector-specific voluntary checklists, and ethics education materials to help organizations apply the principles. One industry team leader warned that the disclosure language in South Korea's principles "could influence the direction of AI-related regulation now being discussed". ANALYSIS That tension — between voluntary norms and the regulatory gravity they may acquire — will be the clearest near-term indicator of whether APAC governance trajectories are converging or fragmenting. For now, the region is running three experiments at once.