Introduction
Mr. Speaker, Sir. I would like to begin by thanking the Prime Minister and Minister for Finance for setting out a clear plan for Singapore in an uncertain world. This Budget speaks not only to today’s pressures — the cost of living, the immediate anxieties of our constituents — but to tomorrow’s transitions.
Over the years, I have had the privilege of working closely with our youths and young workers in my various roles. Their hopes, their aspirations, and their quiet concerns have always stayed true to my heart.
Today, our young Singaporeans are entering the workforce at a moment of profound historical inflection. It is a time of rapid change. Artificial intelligence is reshaping the very definition of a "job”. Economic restructuring is creating pockets of uncertainty, and the traditional pathway from school to work—once a relatively straight, well-lit expressway — might no longer so straightforward.
If we look at the macro data, the picture is indeed reassuring. In Singapore, education and employment outcomes for our youths remain strong. Our students continue to excel consistently in international assessments such as the Programme for International Student Assessment (PISA). Youth unemployment has remained stable in recent years.
But Sir, numbers alone might not always tell the full story.
We hear on the ground worries about the pathways into the workforce. We hear of concerns over entry-level opportunities and career progression. There are reservations about whether hard work would still translate into advancement in a fast-moving economy shaped by technological changes, cost pressures, and global volatility.
As a case in point, this year’s Lunar New Year conversations took on a different texture.
We saw the modern touches, of course. There were the Nano Banana-generated Lunar New Year family portraits and video greetings. We joked about how AI could never replace the traditional Lohei—the human spirit of tossing to prosperity.
But there were also conversations that were far more serious. We talked about the impact of AI on work, especially on entry-level jobs.
I heard concerned parents speak of how much harder it was for their children to get jobs after school. I heard of some young Singaporeans admit quietly, often with a sense of frustration, about sending out hundreds of job applications before they even got a single interview.
And I heard the other side, too. I heard bosses stewing about how they could deploy AI faster in their companies to overcome the manpower crunch. I heard supervisors wondering how they could get their juniors to complement and supervise AI in their daily work.
Everyone is trying to grapple with how AI should work for them.
Our fresh graduates are capable. They are hungry. Yet, many are not fully confident that they are equipped for this new job market.
This is not just anecdotal. Last year, a quarter of youths from our Institutes of Higher Learning (IHL) polled in NTUC’s Youth Employment Outlook expressed worries about finding a job within six months of graduation.
They ask: “Will there be enough entry-level jobs for me?”
When these youths enter the workforce, they do face competition and a narrowing of good entry-level roles. They ask, “Will the skills I’ve acquired be relevant, especially in this age of AI?”
Their concern is also sharpened by technological change. From NTUC’s Survey on Economic Sentiments in 2025, the perception that technology or AI will be the greatest risk to their jobs in the next year was highest among younger workers—19%, compared to the average of 16%. 53% of younger workers, more than any other age group, recognised the need to upskill to stay relevant, even if they could keep their job.
For them, AI is not just a tool; it is an immediate disruptor that is already reshaping some entry-level jobs.
The Broken Rung
What they are worried about is the focus of my speech today. It is a critical structural challenge facing our workforce. It is a challenge that might threaten to change the social compact for an entire generation of Singaporeans in the years to come.
I call it the “Broken Rung”.
In the past, a fresh graduate entered the workforce and learned by doing. The junior accountant drafted working papers. The junior lawyer reviewed documents. The junior coder does code maintenance and fixes minor bugs. These were not glamorous tasks. They were routine. But they were formative.
They were the training ground.
Today, generative AI performs or could perform many of these routine cognitive tasks in seconds. What once took a week now takes an hour. What once required many juniors may now require a somewhat more senior person armed with some AI tools.
Let us be clear: AI does not spell the end of work as we know them today. Many jobs will not be adversely impacted, in fact many might still be boosted by AI.
Research from a 2023/24 National Bureau of Economic Research studying five thousand customer support agents found that AI assistance significantly boosted productivity, especially for less experienced workers. In some cases, a two-month-old employee performed like someone with six months of experience.
This means that the learning curve has been shortened. The barrier to competence has been lowered. The AI did not replace the entry-level worker; it empowered them. It allowed them to punch above their weight almost immediately.
That is not displacement. That is augmentation.
There may also be what economists call a “rebound effect.” If AI allows us to build software faster, we may build more software. If it lowers the cost-of-service delivery, demand may expand. This expands the economic pie.
But here is the uncomfortable truth: even if AI raises overall productivity, it can still shrink traditional entry-level roles.
Why? Because of incentives.
Entry-level workers have historically been hired not because they were immediately productive, but because firms invested in their future productivity. Companies paid a premium today to build capability tomorrow.
If AI can perform the training tasks — the “learning tasks” — faster and cheaper, the short-term economic incentive to hire juniors weakens.
This can be a market failure.
Left entirely to market forces, firms may rationally choose fewer juniors and more AI. In the short term, that boosts margins. In the long term, it hollows out the pipeline of middle-level Singaporean talent.
Compounding this, our higher cost structure and manpower tightness in Singapore have prompted some companies to site teams of knowledge entry-level workers in the region, while retaining the quality control and supervisory roles here. This is understandable as companies seek to control costs while retaining quality.
In isolation, this is not a bad situation for Singapore, as we retain the higher-value work. However, this also means that entry-level jobs might not be available in the same number over time. And this might worsen the hollowing out of the Singaporean middle management.
We risk what I call a “Missing Middle” future: powerful algorithms, a thin layer of senior experts, and too few local managers and specialists in between.
That would not just be an economic problem. It would be a social one.
Our youths are looking at this landscape and asking, "If the AI can do the junior entry-level work, how can I progress further?"
The crux is not whether the job exists, but whether the human can adopt and adapt to AI in his or her work. As Secretary-General Ng Chee Meng and SMS Desmond Tan have shared in their speeches, the nature of the work is changing, but the need for human work remains.
Our economy continues to attract new investments. However, each dollar of new investment now generates fewer jobs than before. This is different from earlier decades, when industrial growth naturally produced many manufacturing jobs and clear entry points. But this is natural and to be expected as the global economic structure changes.
According to economist Dani Rodrik, the future of jobs will lie largely in services. Last year, job creation in Singapore was mainly driven by investments in services, particularly in financial, health, and social services. However, as globalisation has made the outsourcing of services increasingly pervasive, this is also the sector where it is increasingly difficult to keep entry-level roles in the same numbers and scope.
The New Industrial Policy
So, Mr. Speaker Sir, the question is not "How do we stop the AI disruption?" The questions are: "How do we keep entry pathways open while we transform?" “How do we prepare for this possible new future of entry-level work?”
We cannot save the old rung. It is probably gone in some industries or going to over time in others. We should not try to preserve "routine" work in the age of AI.
Instead, the solution to the Broken Rung is not to repair the old one, it is to create a new one.
We need a new industrial policy for entry-level jobs. We must fundamentally redefine what it means to be "entry-level" in the Singaporean economy of the future.
In the past, an entry-level worker’s value was production: write this memo, draft this code, or compile this report.
In the future, an entry-level’s worker value must increasingly be supervision, integration and judgment.
Audit the AI’s output for truth and bias. Integrate systems securely. Provide context and empathy machines cannot replicate.
This is a higher-value role. It is also a harder role because it assumes a level of maturity and skill earlier in one’s career.
And here is where we need a new industrial policy for entry-level jobs to prepare for the new frontier.
Singapore has never been passive in the face of structural change. When we industrialised, we did not wait for jobs to appear — we built the ecosystem. When computerisation emerged, we trained workers. When low-wage stagnation threatened dignity, we implemented the Progressive Wage Model to redesign jobs.
We have similar conviction for the AI era, as expounded by many of our members today. We can put in place a policy framework ahead of the potential change.
Allow me to outline a possible three-pronged approach.
We often admire the apprenticeship systems of Germany and Switzerland. Apprenticeships there are not confined to trades. They extend into banking, insurance and advanced services. The state and industry partner to standardise training so that apprentices contribute productively from the start.
Why should we not adapt this approach to our knowledge economy?
We can incentivise firms to convert traditional execution roles into structured apprenticeships with core elements of “AI oversight” or “AI integration”. For example, if a law firm hires a graduate to audit AI-generated contracts — not just draft from scratch — that firm may face an initial efficiency gap. The AI is faster.
Government can co-fund part of that gap for a defined period, de-risking the hiring decision while building national capability.
Over time, we can formalise AI oversight as a recognised skill within industry frameworks — portable, certified, and valued across employers.
It is an investment in the next generation of Singaporean middle managers.
We will also need to identify "Queen Bees"—industry leaders to lead such transformation in the workforce.
Denmark offers a promising initiative. Through its national AI Skills Pact, it has committed to strengthen AI capabilities among 1 million citizens by 2028, positioning itself as a leading human-centred AI nation. The approach is deliberate: joint flagship projects, employer-led commitments and coordinated capability-building across sectors. The state sets direction: industry and unions lead execution. We can think about this as an investment for our future generation of middle managers.
Second, equip our students for semi-senior entry responsibilities from Day One of their working lives.
We must accept that routine tasks are no longer guaranteed training grounds.
AI literacy cannot be optional. It must be foundational. Every graduate should be proficient not only in theory but in practical AI usage — prompt engineering could be a thing of the past, output verification, ethical evaluation, and knowing when human judgment must override machine recommendations.
But technical skill alone is insufficient.
As machines take on routine tasks, uniquely human capabilities become more valuable: communication, partnership, leadership, and ethical reasoning.
Initiatives such as NTUC’s Youth EXCEL programme, one of the youth programmes, and youth-led policy platforms are important because they do more than train for employment. They cultivate agency. They signal to young Singaporeans that they are not passive recipients of change, but participants in shaping it.
Allow me to share an example. I looked at our latest cohort of NTUC Starter Awards participants. The winning team presented on the idea of cross-sector internships. For example, Valerie, one of the team members and a student at Temasek Polytechnic, the experience showed her how youth-led insights can meaningfully contribute to broader policy discussions and industry changes.
Our young people have the ideas. We need to give them the platform.
Third, accelerate reskilling timelines.
Under current arrangements, SkillsFuture credits are accessible later in a worker’s career. But AI compresses learning and skill cycles. A diploma holder entering the workforce at, say, 21-year-old may find within a few years that additional credentials are necessary. In fact, many of our current graduates might already need to complement their academic training while in school to be ready for the new workplace demands.
We should examine whether earlier access to training support might be needed in this new environment.
At the same time, we must expand place-and-train opportunities and career bridges models. Reskilling is hardest when there are bills to pay. Structured pathways that combine employment with training reduce risk and increase participation.
Conclusion
Mr Speaker Sir, some may ask: can Singapore manage the sweeping changes AI will bring?
Our history says yes.
Each time our economy evolved, we did not retreat. We responded with policy creativity and tripartite partnership.
The question before us is not whether AI will transform work. It will.
The question is whether that transformation strengthens or weakens the Singaporean promise.
The "Broken Rung" is a structural shift, but it need not be a structural failure.
If we do not innovate ahead of time, we risk a "Missing Middle" crisis—a hollowed-out workforce. We risk discouraging our youths and weakening our social compact.
But if we act now to recreate the entry-level job, we can turn this disruption into an opportunity. We can build a workforce where every entry-level worker is an architect of their own future, supported by technology rather than replaced by it.
We ensure that the Singaporean promise remains unbroken for the next generation—that if you work hard and master your craft, there is a ladder for you to climb. NTUC will do its utmost to support our workers.
Let us build this new ladder together.
Mr Speaker, Sir, I support the Budget.
