Singapore's AI Talent Crisis: The Build vs Use Dilemma

Date: March 13, 2026

Singapore has committed over S$1 billion to become a global hub for artificial intelligence. But one of the architects of that effort is sounding the alarm: the nation's approach to AI readiness may not be sufficient.

The Build vs. Use Gap

Leslie Teo, senior director at AI Singapore, the national AI research and development programme established in 2017, warns that the current approach risks focusing on producing certified AI users when the nation desperately needs more AI builders.

"Junior employees are cheap. AI is cheaper, though," Teo said in an interview on Thursday (March 12). This stark observation highlights the growing challenge as companies increasingly adopt AI technology while simultaneously cutting back on junior hiring.

The implications are significant. As businesses automate more tasks, the traditional pathway for young graduates to gain experience and climb the career ladder is shrinking. The share of fresh graduates securing full-time permanent jobs fell to 74.4% in 2025, down from 79.4% the year before, according to the latest annual graduate employment survey by Singapore's universities.

Training at the Speed of AI

Teo, whose career has spanned global financial institutions and the tech industry—including roles at the International Monetary Fund, the Monetary Authority of Singapore and Grab Holdings—believes the Singapore government will need to step in and treat early-career training as a public good as firms reduce the training they once provided new hires.

"The thing about AI is what you know today and what you know tomorrow can dramatically be opposite each other," Teo explained. This rapid pace of change exposes a fundamental flaw in Singapore's current training infrastructure: the time it takes to design, approve and roll out formal training programmes means they can be years out of date by the time a curriculum is approved.

At the moment, the country's main tool to address this is SkillsFuture, the national training credits programme that subsidises courses for citizens throughout their careers. About 606,000 individuals took part in supported training programmes in 2025. While participation is high, the issue lies in speed.

National Sovereignty in the AI Era

For Teo, the stakes go beyond the job market. Countries that rely entirely on AI systems developed elsewhere risk having little influence over how the technology evolves or whose interests it serves.

"There are certain decisions about technology and building that if you're not at the table, you cannot say so," Teo cautioned. This sentiment echoes growing concerns among policymakers worldwide about technological sovereignty in an era dominated by American and Chinese AI systems.

Singapore is trying to secure that seat at the table. Its AI Singapore programme has developed Sea-LION, a large language model designed for South-east Asia and already used by regional companies including GoTo Group. The government says it is narrowing the gap by bringing in technical talent and training workers in AI.

Government Response

A tech and AI track added to the Overseas Networks & Expertise Pass under Singapore's 2026 budget is expected to attract foreign specialists, alongside programmes to help industries adopt the technology.

"You bring in talent, you groom existing talent across sectors, and you make it quite fertile for people to adopt these use cases quickly," said Alvin Tan, minister of state for trade and industry.

The Path Forward

Still, the question facing Singapore is ultimately about capability rather than participation: whether the country can develop enough people capable of building AI systems, rather than simply train workers to use them.

The challenge requires rethinking how Singapore approaches AI education—from creating faster curriculum cycles to investing in foundational AI research capabilities. As the technology continues to evolve at breakneck speed, the nation must decide whether to be a consumer of AI or a creator of it.