Singapore’s AI pivot will require a very different playbook

The Economic Strategy Review (ESR), whose 32 final recommendations were unveiled on May 13, is a bold blueprint for economic transformation. Launched amid intensifying geopolitical turbulence and accelerating AI disruption, it sets out an ambitious agenda for keeping the Singapore economy competitive, connected and relevant to the times.

Its goal is to secure growth at the higher end of the 2 to 3 per cent range over the next decade, while ensuring that it translates into good jobs and rising incomes, not just aggregate prosperity; the ESR is right that good jobs can no longer be assumed to come automatically with growth. But ambition on paper must be translated into transformation on the ground. Delivering on the ESR’s agenda will require at least five policy leaps that go well beyond what Singapore has attempted before.

1. Attract young companies differently

Singapore has been a world-class destination for established multinational corporations for six decades. But that game is getting harder. With competition for their investments intensifying, MNCs are distributing operations across more locations. Besides, they are less likely to bring transformative new capabilities than dynamic, younger firms might. Catching high-potential companies before they become global giants is where the real long-term value lies.

The ESR recognises this, calling for a more vibrant start-up ecosystem with stronger support for high-potential firms, and positioning Singapore as the place to develop, test, deploy and scale impactful AI solutions. But the policy machinery needs to match the ambition.

The capital problem is real. Singapore has built excellent infrastructure for early stage start-ups – Startup SG, SEEDS Capital and a dense network of accelerators have made it easy to get a company off the ground. Budget 2026 added a $1 billion expansion of the Startup SG Equity scheme to support growth-stage deep tech companies, which is a meaningful step.

But the gap between a well-funded early stage start-up and a company ready to attract global institutional investors remains wide. Singapore needs top-tier global venture capital (VC) firms such as Sequoia, Andreessen Horowitz and Insight Partners to establish investment teams in Singapore, not just regional offices. That requires deliberate courtship.

One incentive could be a guarantee of co-investment by a Singapore sovereign wealth fund or the state investor for any qualifying deal sourced by the Singapore VC team. Another could be to create a credible exit pathway for VCs through the SGX-Nasdaq dual listing bridge announced in Budget 2026, coupled with liquidity support from government funds for a minimum period to ensure that the listing does not become illiquid.

In addition, Singapore must leverage its regulatory advantages. Many high-potential young companies are being built in areas where there is much regulatory uncertainty: in AI and autonomous systems, biotech and genomics, fintech and digital assets, autonomous vehicles and space technology. The regulatory sandbox model that the Monetary Authority of Singapore has built for fintech companies to test products before full deployment should be extended and deepened across all cutting-edge industries, not just fintech.

Singapore’s small size is also an advantage in that it can be used as a living laboratory, with the Government piloting new technologies in real urban conditions — as it has done with autonomous vehicles on public roads and cashless payments. That would be a compelling attraction for companies seeking real-world validation data to scale globally.

If the Government offers itself as a first customer for tested technologies, that also would help young companies prove their products and become more easily fundable and scalable globally.

2. Fix the AI talent shortage

General Assembly’s State of Tech Talent 2026 report found that 95 per cent of employers in Singapore face challenges hiring for tech roles, with 58 per cent identifying roles in data analytics and data science as the hardest positions to fill. Globally, AI talent demand exceeds supply by more than three to one, with machine learning engineers, large language model developers, machine learning operations specialists, AI product managers, AI ethics officers and cybersecurity specialists among the most acutely short-skilled categories.

Singapore cannot train its way out of this shortage quickly enough. The pipeline from universities takes years; the technology is moving in months. So, Singapore needs to significantly increase its import of senior AI specialists – people capable of both operating on the frontier and training others locally. Mid-level data and analytics roles can be filled locally with upskilling, while some functions can be outsourced. But the senior layer needs to come from outside.

The ESR acknowledges the talent challenge. What it must also acknowledge is that solving it requires a degree of openness on foreign hiring that will test political comfort zones.

3. Rebuild SkillsFuture for wholesale career reinvention

SkillsFuture has done something genuinely valuable: it has normalised the idea that learning does not end with formal education. But it was designed for a gentler disruption – one where workers needed occasional top-ups to existing skills, not wholesale reinvention. In the era of AI, that model is inadequate.

Its fundamental flaw is that it offers skilling that is mostly episodic in nature. It provides credits for discrete, usually one-off courses, without building sustained relationships between learners and providers, tracking what skills have been acquired, or mapping a pathway to where the learner needs to go. The course marketplace is broad but full of low-value certifications that employers do not recognise and that rarely translate into employment outcomes.

AI displacement requires something different. A paralegal whose core tasks have been automated does not need a weekend workshop – she needs 12 to 18 months of structured deep retraining, with income support, mentorship and active job placement at the end.

That is a programme, not a course. The ESR’s Thrust 7 – empowering workers to learn for life – is a step in the right direction, but making lifelong learning a practical reality requires making training more structured and blending it with work, not simply expanding the menu of courses available.

A reformed SkillsFuture – merged with Workforce Singapore into a single statutory board combining reskilling and reemployment functions – needs to do three things: provide sustained, sector-specific pathways rather than isolated courses; collect and act on data about what skills gaps remain after training; and guarantee active job placement support, not just training access. Reskilling without re-employment is a half-measure.

4. Replace the job seeker safety net with wage insurance

The ESR is explicit that Singapore must protect workers, not jobs. But the mechanisms available are mismatched to the problem.

The SkillsFuture Jobseeker Support scheme, which provides up to $6,000 over six months to involuntarily unemployed workers — amounting to $1,000 a month – is not fit for purpose in the AI era. It was designed for a labour market where displacement was concentrated among lower-income workers. But in the AI era, it will increasingly be professionals – such as lawyers, analysts, accountants and coders – whose core tasks will be automated.

For a professional earning $8,000 to $15,000 a month, this replaces a small fraction of lost income. It would not cover a mortgage or basic living costs. Faced with that financial reality, displaced professionals will not retrain purposefully – they will likely accept the first available role and take whatever income cut it requires, losing out on a meaningful career pivot. Six months is also too short a runway for serious retraining: moving from financial analysis to data science, or from legal work to AI governance, takes at least a year to reach the level of proficiency needed to attract a competitive salary.

The ESR recognises that it must address income loss. Wage insurance would be an appropriate instrument. Under a well-designed scheme, a displaced professional who takes a lower-paying job during transition would receive a government supplement – say, 50 per cent of the income gap between the old and new salary – for up to two years.

This would cushion the financial fall without removing the incentive to re-enter the workforce quickly, and provide the runway needed for genuine retraining rather than panic re-employment. It is not a welfare measure like standard unemployment insurance, which Singapore has rejected before because of the danger of moral hazard. Rather, it is a scheme that makes career transition economically viable.

5. Tax the productivity dividend to fund the transition

Social safety nets, wage insurance and a rebuilt SkillsFuture will be expensive. The question of how to fund them without undermining Singapore’s attractiveness as an investment destination is the central fiscal challenge of the AI transition. If AI generates large productivity gains they will accrue mainly to capital owners and highly skilled workers while displacing lower-and mid-skilled workers. As a result, inequality will widen and social trust will erode.

The answer to funding is not higher corporate tax rates, which are constrained by the Organisation for Economic Cooperation and Development global minimum and Singapore’s competitive imperatives. The more targeted option is a modest levy on revenues generated by large AI and digital platform businesses operating in Singapore – applying only above a high global revenue threshold, say $750 million, to exempt the high-growth younger firms Singapore is trying to attract; levied on local digital revenues rather than profits to reduce avoidance; and set at a modest 2 to 3 per cent, generating potentially several hundred million dollars annually without creating a significant cost to the firms it targets.

Critically, the revenue should be hypothecated directly to transition funds. Companies subject to the levy are more likely to accept it if they can see it funding the workforce transition from which their own talent pipeline will ultimately benefit.

Implementation: the hard part

The ESR was convened because Singapore’s leaders recognise that the rules that allowed Singapore to prosper have changed fundamentally – that growth will be harder, that good jobs are no longer guaranteed by growth alone, and that Singapore must move faster and take calculated risks. They are right on all counts.

But implementation must match this ambition. Singapore has managed structural economic transitions before – from entrepot trade to manufacturing, to advanced manufacturing and sophisticated services. The AI transition is harder, faster and more disruptive than anything that came before. Meeting it will require not just a bold report but the willingness to redesign some of the most familiar features of Singapore’s economic model. The ESR has shown the direction. Now comes the harder part.