Back
Singapore’s job market looks healthy. Employment is rising, unemployment sits at 2 per cent, and nearly 1.5 job vacancies still exist for every jobseeker.
Yet, ask any recent graduate and you will hear the same complaint: Getting started feels harder than the numbers suggest.
Skill requirements in Singapore are now changing roughly 2.4 times faster than in the US, and entry-level roles churn fastest of all. The issue is not a shortage of jobs; the bar for securing one has quietly moved higher.
Artificial intelligence is often blamed for that, but Singapore’s graduate labour market had already tightened before the technology became mainstream.
AI has simply arrived on top of that trend, giving employers another reason to maintain higher expectations.
The escalator has stalled
For much of the 2000s and early 2010s, Singapore’s entry-level roles followed a predictable escalator: A graduate entered a junior role, did the routine work, built judgment through repetition, and in time moved up into management, opening up her old role for the next graduate behind her.
The Institute for Adult Learning’s (IAL) analysis of Singapore’s labour force data shows that pathway slowed around 2014. The share of managers holding a degree rose from about 20 per cent in 1991 to roughly 65 per cent by 2014, before plateauing.
Instead, graduates who entered the workforce after 2014 increasingly took on associate-professional, technical and clerical roles that historically did not require degrees.
However, the wage premium associated with these roles was smaller than it had been for earlier graduate cohorts. By 2019, workers in these occupations typically earned slightly below Singapore’s median income.
Evidence points to a structural shift
ChatGPT, the first popular generative-AI tool (GenAI), was released in 2022. In other words, AI did not cause the labour market squeeze, but it is tightening it.
Joint analysis by IAL and the Burning Glass Institute (BGI) of about 15.3 million job postings found entry-level roles declining as a share of vacancies in both Singapore and the US, after GenAI emerged.
Postings requiring fewer than three years’ experience fell most sharply in applied fields such as engineering and software development.
Consider a junior data analyst. A few years ago, her first months would have been spent cleaning data and checking figures.
Today, AI does that work in seconds.
Her value lies in judging whether the output is right, identifying what matters and explaining it to a client. Skills once learnt on the job are increasingly expected from day one.
BGI calls this an “expertise upheaval”: If AI performs the foundational work that once helped people develop professional judgment, organisations must rethink the way new employees acquire it.
IAL’s wider research on AI adoption by employers between 2019 and 2023 points the same way. It found most organisations expected existing teams to do more with AI rather than significantly expand professional headcount.
Meanwhile, its latest report, Navigating AI-Driven Skill Change, found entry-level roles experienced the fastest skill churn, meaning that previously relevant expertise is quickly becoming obsolete.
In particular, between 2023 and 2025, job postings for knowledge workers listed fewer explicit skills than those put up before GenAI, suggesting employers increasingly assume a higher baseline of capability.
Together, these findings suggest the rising bar for entry-level hiring reflects the expectations that businesses have about hiring talent, not a technological inevitability. That is, AI accelerated an already existing structural issue.
Singapore is not alone. The same pre-AI pattern of increasingly weaker entry-level hiring shows up in the US and the UK, indicating a shift societies are learning to manage, rather than a uniquely Singaporean challenge.
Why some graduates are thriving
The challenges are unevenly tackled. Graduates with internships, industry-linked programmes or strong networks remain better positioned to demonstrate work readiness.
Others, with similar potential but fewer opportunities, face a circular problem: The experience now expected for entry-level roles is the experience those roles once provided.
This conundrum is compounded by the wider economy. Three in four small and medium-sized enterprises (SMEs) in Singapore have weak business strategies and compete on cost rather than value, so they do not prioritise employee development and training.
Without a stronger middle tier of companies creating quality professional opportunities, those who miss traditional graduate programmes have fewer pathways to build the experience needed by today’s labour force.
AI plays a role here too. To be competitive in today’s entry-level jobs market, graduates should treat AI fluency as necessary, but not sufficient.
AI supervision capabilities – a step above fluency – already attract a premium.
Singapore salaries for such roles have reportedly risen 15 to 25 per cent in the past year, with some salaries for fresh hires starting at between S$70,000 and S$90,000 annually.
While these jobs remain a small share of entry-level hiring, they show that it is possible for fresh graduates to secure such coveted roles if they demonstrate the skills – not just credentials – that employers are looking for.
In today’s AI-driven workforce, such skills include being able to show, during an interview or early on the job, that they can catch what a model missed.
That is worth coaching into junior hires, and worth sharing with anyone wondering why today’s job market feels harder than the headlines suggest.
Employers face a tough question: Should some “learning work” be deliberately preserved, even where AI could do it faster?
A graduate programme optimised purely for output will keep producing people who can operate AI, but never learn to judge it.
Structured mentoring and rotational postings cost more upfront than letting AI handle the first draft, but they are the only proven way to grow the judgment a company will need in 10 years.
Educators and policymakers have levers too.
Universities and polytechnics can ride on the experience of work-study programmes to further expand work-integrated learning, so that every graduate has the opportunity to demonstrate value-creation and strengthen AI supervision skills.
Policymakers can extend existing schemes such as the TechSkills Accelerator and SkillsFuture’s Workforce Development Grant to incentivise companies to preserve structured-learning roles for fresh hires, while industry associations can pool internship placements across SMEs too small to run graduate programmes alone.
AI will continue reshaping work, but the bar for entry-level employment began rising long before the technology became part of everyday business.
However, it can be brought within reach of fresh graduates again, by the same means it rose: the choices employers, educators and policymakers make.
Singapore has rebuilt its ladder of opportunity before. There is no reason it cannot do so again.
The commentary by Ms Kyla Lee, analyst at the Centre for Skills-First Practices, Institute for Adult Learning was first published in The Business Times on 22 July 2026.