Singapore is pursuing artificial intelligence policy on two parallel tracks, throwing its support behind a new international call for stronger safeguards on frontier AI systems while continuing to expand the technology’s use across its own public sector.
Prime Minister Lawrence Wong said on September 22 that Singapore supports an international statement urging tighter safety measures around the most advanced AI systems. The initiative was launched by Norwegian Prime Minister Jonas Gahr Støre and Finnish President Alexander Stubb, and has now been endorsed by Wong along with 21 other world leaders. The statement calls for companies developing frontier AI to maintain transparent safety protocols, carry out mandatory testing before deployment, and submit their systems to independent evaluation.
Push for Coordinated International Standards
Beyond company-level safeguards, the statement asks governments and regional organizations to work together on common safety standards, share reports when serious AI safety incidents occur, and widen access to scientific expertise and trusted evaluation resources. It also calls on United Nations member states to look into establishing an international institution that could set standards for frontier AI, support verification efforts, and bring together states when certain AI capability thresholds are reached. Wong said that because AI-related risks are not confined within national borders, meaningful safeguards depend on cooperation between countries.
AI Already Widely Used Inside Government
Domestically, AI adoption within Singapore’s public administration is already substantial. Digital Development and Information Minister Josephine Teo said close to 70 percent of the country’s public officers now use AI tools on a regular basis, including for tasks such as analyzing information and drafting documents. Government agencies are also examining how AI agents could reshape existing workflows, while ensuring that humans remain accountable for significant decisions.
AI systems are being deployed across a range of public functions. Machine-learning tools are used to direct public feedback to the appropriate government agencies, while AI assists officers at border checkpoints in spotting irregularities among large volumes of cargo. Similar technology is being applied to counter scams, helping identify patterns associated with phishing and impersonation attempts. Teo said maintaining public trust is critical to continued adoption of these tools, especially since AI can also be used to make scams and cyberattacks more sophisticated.
Expanding Access Through Regional Language Tools
Singapore is also working to broaden participation in AI development beyond its own borders. Through an initiative called SEA-LION — Southeast Asian Languages in One Network — the country is developing open AI models tailored to Southeast Asian languages and cultural contexts. Separately, AI Singapore is partnering with international organizations and development partners on Project ATLAS, an effort to build AI-ready datasets for languages that are otherwise underrepresented in AI development. Singapore is also sharing digital resources and expertise through the Digital Pillar of the Forum of Small States and the Singapore Digital Gateway.
Taken together, Singapore’s latest steps combine engagement on international AI safety cooperation with continued deployment of the technology across its own government services, an approach officials say is meant to balance broader adoption with oversight and public trust.
This article references reporting from:
