SINGAPORE, Jan 24, 2026, 21:19 SGT
- Singapore will invest more than S$1 billion in public artificial intelligence research through 2030.
- The plan targets basic research, industry use-cases and a wider talent pipeline from schools to faculty.
- Officials flagged the power and water costs of running advanced AI as Singapore manages data centre growth.
Singapore will invest more than S$1 billion (about $780 million) in public artificial intelligence (AI) research through 2030, deepening a push to build national capabilities and to get more companies using the technology, the government said on Saturday. Reuters
The funding lands as AI tools move from demos to deployment, and as countries scramble for the computing power and talent needed to train and run the models behind them. The race is still being shaped mostly by U.S. and Chinese tech groups, and smaller economies are trying to pick their spots.
Singapore is framing this as a research play, not just a procurement plan. National AI Strategy 2.0, announced in 2023, set a goal of more than tripling the number of AI practitioners to 15,000, and officials have pitched the city-state as a place where leading AI builders can base teams.
The Ministry of Digital Development and Information said it will steer money to priority areas such as responsible and resource-efficient AI, while building a pipeline from pre-university students to faculty. It also wants more tools that industries can apply, not just papers.
The five-year National AI Research and Development Plan runs from 2025 to 2030 and was announced by Digital Development and Information Minister Josephine Teo at the Singapore AI Research Week 2026 gala dinner, the ministry said. Teo warned the computing bill is rising. “For example, AI training and inference remain extremely resource-intensive. Their draw on energy and water cannot be ignored,” she said. (Training is the compute-heavy process of building a model; inference is using it to generate results after it is built.) Channelnewsasia
A significant slice of the money will go into AI research centres of excellence, hosted in public research institutions and staffed by local and international researchers, Teo said. Business Times reported the new centres would sit alongside an existing network of more than 60 AI centres of excellence, but be fewer in number and backed by larger sums per centre.
The research agenda includes work on responsible AI — systems designed with safeguards against misuse — and on ways to reduce AI’s reliance on data and computing resources. Officials have also pointed to “general-purpose AI”, meaning models intended to handle multiple tasks across different domains.
Applied research will focus on closing the gap between lab work and deployment, including projects driven by national research and enterprise programmes, the ministry said. Authorities will work with industry partners such as Changi Airport Group and Sembcorp to build core AI engineering capabilities and test use-cases at speed and scale, Business Times reported.
Singapore is trying not to compete head-on with global leaders building the biggest “foundation models” — large, general AI systems trained on broad data that can be adapted for many jobs. Business Times said Singapore instead wants to build and adapt foundational capabilities, then tailor models for regional languages and security needs.
That approach echoes work by AI Singapore, a national programme that has backed Sea-Lion, an open-source large language model for Southeast Asian languages. Reuters reported companies including Indonesia’s GoTo have adopted the model.
Reuters also reported an updated Sea-Lion model released in October 2025 was built on top of Qwen, a foundation model released by China’s Alibaba, with improved support for Burmese, Filipino, Indonesian, Malay, Tamil, Thai and Vietnamese. Singapore has separately set aside S$500 million for high-performance computing resources, the powerful computer capacity typically used to train and run AI systems.
But the funding does not magic away physical limits. Teo has pointed to the strain AI places on electricity and water, and the pressure to manage data centre expansion against Singapore’s net-zero commitments. If computing capacity tightens or industry uptake lags, the research push could take longer to translate into products and wider adoption.
Teo said she wanted the ecosystem to stay diverse and tight, not siloed. “We believe good outcomes will emerge out of a vibrant, diverse yet close-knit research ecosystem,” she said. Com
