AI is tightening Singapore's entry-level job market — but the squeeze started years before ChatGPT

Singapore's headline employment numbers look as healthy as they have in years — unemployment at 2 per cent, nearly 1.5 vacancies for every jobseeker, and employment ticking upward across most sectors. But beneath that surface-level picture, something structural is shifting beneath the feet of recent graduates and early-career professionals. The bar for landing a first job has quietly moved higher, and new data suggests the trend predates the AI boom by nearly a decade.

Skill requirements in Singapore are now changing roughly 2.4 times faster than in the United States, and entry-level roles are churning fastest of all, according to a joint analysis by the Institute for Adult Learning (IAL) and the Burning Glass Institute (BGI) that examined 15.3 million job postings across both markets. The IAL's workforce data reveals that the traditional graduate escalator — join at a junior role, do routine work, build judgment, rise into management — began slowing around 2014. By that year the share of managers holding degrees had plateaued at roughly 65 per cent, after climbing steadily from 20 per cent in 1991. Graduates entering the workforce after 2014 increasingly landed in associate-professional and technical roles that historically did not require degrees, and by 2019 those occupations paid slightly below Singapore's median income.

Generative AI did not create this pressure — ChatGPT launched in 2022, years after the trend had taken hold — but it is accelerating it. Postings requiring fewer than three years of experience fell most sharply in fields like engineering and software development, precisely the areas where AI tools can now perform the foundational work that once helped junior hires build expertise on the job. The BGI describes the dynamic as an "expertise upheaval": if a model can clean data, draft code, or produce a first-pass analysis in seconds, the junior role that existed to do those tasks necessarily shrinks. What remains is the higher-order work — judgment, interpretation, stakeholder communication — that employers now expect from day one rather than teaching it on the job.

The knock-on effects are visible in hiring patterns and compensation data. Salaries for AI-adjacent roles in Singapore have risen 15 to 25 per cent over the past year, with some fresh hires commanding S$70,000 to S$90,000 annually. But those premium figures mask a broader challenge: three in four small and medium-sized enterprises in Singapore have weak business strategies and compete on cost rather than value, leaving them ill-equipped to train and absorb graduates into an AI-shaped labour market.

Why it matters for Singapore: The data makes clear that AI is not the root cause of the tightening — it is amplifying a structural shift that had already begun in the mid-2010s. That distinction matters for policy. Extending existing programmes like the TechSkills Accelerator and SkillsFuture's Workforce Development Grant can help, but the deeper takeaway is that Singapore needs to rebuild the ladder that quietly stalled a decade ago. As the BT piece puts it, this country has rebuilt its ladder of opportunity before. With AI now raising the rungs faster than ever, doing so again is less a choice than a necessity.