Singapore Shifts AI Strategy From Experiments to Real-World Deployment
Singapore is entering a new phase in its artificial intelligence journey, moving away from small-scale pilots and proof-of-concept projects toward practical AI deployment across businesses, public services and key industries.
The country’s updated National AI Strategy, refreshed in May, signals a clear change in direction. Rather than competing only in the race to build the largest AI models, Singapore is focusing on how artificial intelligence can be adopted reliably, securely and at scale. The strategy is being guided by the newly formed National AI Council under Prime Minister Lawrence Wong, with a continued emphasis on using AI for public good.
A major part of this push is enterprise adoption. Through the National AI Impact Programme, Singapore aims to help 10,000 companies adopt AI over the next three years. This comes as AI usage among small and medium-sized enterprises in Singapore has already grown sharply, rising from 4.2% in 2023 to 14.5% in 2024.
For Singapore, the next competitive advantage in AI is not just about building bigger models. It is about helping organizations integrate AI into daily operations in ways that deliver measurable results.
Government Chief AI Officer Dr He Ruimin explained that the AI landscape has changed quickly since 2023. The earlier focus was largely on whether generative AI models could produce useful text. Today, the conversation has shifted to multimodal systems, reasoning models, AI agents and early forms of embodied AI.
In his view, the central question is no longer whether AI can perform useful tasks. It is whether companies and public agencies can deploy AI meaningfully and dependably inside their operating systems.
That focus is shaping Singapore’s priorities across four major sectors: connectivity, advanced manufacturing, healthcare and finance. Together, these industries contribute more than 40% of the country’s GDP, making them critical to long-term economic growth.
Singapore wants AI to become part of the real machinery of these sectors. That could mean improving airport coordination, optimizing factory workflows, supporting medical diagnostics, or strengthening financial crime detection. The goal is to solve complex problems that matter, while attracting global AI companies, researchers and talent.
Dr He described the ambition as intentional. Difficult problems, he noted, tend to attract the people capable of solving them.
This approach is already being tested through physical AI projects. Singapore’s Infocomm Media Development Authority, JTC and the Singapore Institute of Technology have partnered with eight industry players to explore and deploy physical AI systems in the Punggol Digital District. These efforts reflect Singapore’s desire to move AI from software demonstrations into real-world environments.
Another important part of Singapore’s AI strategy is infrastructure resilience.
Because Singapore has limited land, energy and water resources, it is not trying to host every AI workload within its borders. Instead, the country is building a hybrid AI infrastructure model that combines local capacity, regional connectivity and selective use of overseas computing resources.
This strategy recognizes that different AI workloads have different needs. Some may require low latency and strong local security controls, while others may be better suited to overseas data centers based on cost, scale or regulatory requirements.
According to Dr He, Singapore’s priority is to maintain the infrastructure, technical capabilities and connectivity needed to remain a trusted and resilient AI hub. Companies should then decide where to place workloads based on practical factors such as price, speed, security and compliance.
Singapore is also strengthening its position through partnerships with major AI and technology players. OpenAI launched an initiative for Singapore in May with a commitment of more than SG$300 million, equivalent to about US$234 million. The country has also expanded collaborations with Google and Nvidia in applied AI, research and development.
Beyond infrastructure and adoption, Singapore is placing strong emphasis on AI governance.
Its Model AI Governance Framework for Agentic AI, first introduced in January and updated in May, offers guidance on risk limits, human responsibility, technical safeguards and transparency. The updated version includes feedback from more than 60 organizations and draws from more than 10 real-world deployment examples.
This governance framework is especially important as AI systems evolve from simply generating responses to taking actions on behalf of users or organizations. When AI agents can execute tasks, make recommendations or trigger decisions, oversight becomes essential.
Dr He warned that an AI agent taking the wrong action can commit an organization to real consequences. That makes governance not a barrier to innovation, but a foundation for safe and trusted adoption.
Singapore’s AI strategy is therefore not centered on owning the world’s biggest model. Its bet is that the future of AI leadership will belong to places that can deploy advanced AI safely, efficiently and at scale.
By 2030, Singapore wants to be known not just as a country that talks about AI, but as one that deploys it well.
