Singapore’s workers brace for AI disruption: ‘it’s inevitable’

When Joshua Tan turned up to Chinese New Year dinner last month, he hadn’t expected to spend the evening defending his career. He was 27 and freshly employed as a junior software engineer in Singapore. The questions from relatives were pointed: is your job safe? Can’t the computer just do it? It was an awkward conversation, equal parts interrogation and familial concern. But Tan knew the anxiety behind the questions was real. Not so long ago, he and his classmates had been so sought-after by employers that many juggled two internships at once. Now, the question that hangs over every conversation in the industry is simple: what happens when the code writes itself? At the moment, artificial intelligence was being used “more like a steroid booster”, Tan said. It handled the grunt work, helping engineers avoid putting in too much overtime. But in future, Tan predicts that the headcount at his company will go down as AI gets better at the tasks more junior employees used to do. “Like any tool, it’s inevitable,” he told This Week in Asia. That sense of inevitability is becoming harder to ignore. In February, Prime Minister Lawrence Wong addressed the nation’s anxieties head-on in his budget statement, acknowledging the worries many workers had been feeling while insisting that Singapore had no choice but to harness AI to stay competitive. His government has unveiled a sweeping package of measures: a new National AI Council, chaired by Wong himself; a National AI Impact Programme aiming to train 100,000 workers to become “AI bilingual”; support for 10,000 enterprises to deepen AI adoption; and, from the second half of this year, six months of free access to premium AI tools for Singaporeans who enrol in selected AI courses. AI is projected to displace 92 million jobs globally by 2030 while creating 170 million new ones, according to the World Economic Forum’s Future of Jobs Report 2025. The net number looks positive. But whether the people on the wrong side of that ledger find their way to the other is an entirely different question. The missing rung Tan uses AI to improve the code he writes, which a senior engineer then reviews. The process works, for now. He is already thinking about what comes next, however – and what gets lost along the way “When I see my managers and seniors, they are quick to pick out things that both the AI and I missed,” he said. “There is a skill set that needs to be developed through experience.” But what happens if fewer junior engineers are hired because AI is doing all the entry-level work? Such a hollowing out of the talent pipeline would not only be bad for the future of the industry, it risks creating a generation that never learns to work independently of AI. This fear, that the first rung of the career ladder is being sawed off, was raised by members of parliament during Singapore’s latest budget debate – alongside concerns that older workers would find it harder to re-enter the workforce and that companies might use AI as a reason to cut training budgets rather than expand them. “We are alert to these risks, and we will act early to prevent such outcomes from taking hold in Singapore,” Wong said. “We will invest more deliberately and more systematically in our people.” In addition to the National AI Council, Singapore also plans to launch a Champions of AI programme aimed at helping companies adopt the technology through tailored support. Not everyone has been waiting for the government to move, however. Some Singaporeans, hit with redundancy or unsettled by the pace of change, are taking matters into their own hands. Priya Rath, 50, enrolled in a part-time master’s programme in AI and innovation at the National University of Singapore (NUS) two years ago after losing her job in the agricultural sector. “I decided to take up the course to upskill myself and stay ahead of the AI curve as it transforms all industries,” she said. Rath said the city state’s SkillsFuture programme, which promotes lifelong learning, needed to provide a clear, dedicated service to match jobseekers with opportunities. “We were left wondering if it was worth it in the end, especially if you were trying to pivot into a new sector,” she said of her experience with that scheme. Those who do not learn how AI works will fall behind Edmund Han, 50, marine engineer Edmund Han, also 50 and a marine engineer, enrolled in the same NUS programme for different reasons. For him, too many people were talking about AI without truly understanding it. “It gave me a better understanding of what AI is, what the limitations are today and some of those areas where AI is utilised and where it still needs to be complemented by human beings in the loop,” he said. Rather than focus solely on technical training, the course offered clarity on the technology’s use. “To me it’s a no-brainer,” Han added. “Those who do not learn how AI works will fall behind.” ‘Lights out’? Ori Sasson, director of cyber-intelligence firm S2T, recently found himself with a dilemma that employers will increasingly recognise. He had wanted to hire a fresh graduate for a junior associate role, but then he crunched the numbers. He estimated that an AI chatbot could handle about 60 per cent of the position’s tasks. And in the end, he didn’t make the hire. The role did not cease to exist, he said. “It’s just that people using these tools are more productive, which means less demand for certain routine roles.” Sasson is careful to make a distinction. The remaining 40 per cent of the role – judgment, context, knowing when the AI is wrong – is where human value lies. But he argues it is precisely these skills that universities are failing to cultivate. “We need to ask ourselves what school is supposed to give you,” Sasson said. His answer? “Judgment, taste, ethical reasoning and the ability to ask the question.” In other words, qualities that large language models (LLM) such as ChatGPT, Claude and Gemini largely lack. “The LLM can only answer well if you know how to ask it well,” he said. Sasson has no patience for universities that respond to AI by banning it outright, as he says doing so does not prepare students for the real world. Jess O’Reilly, general manager for Southeast Asia at human capital firm Workday, maps out AI disruption by the sector. She said industries built on “routine cognitive” tasks – customer service, data entry, traditional accounting and audit functions – would see the greatest impact. In the financial sector, O’Reilly highlighted a “move towards continuous, real-time auditing and ‘lights out finance’, which was humanly impossible before”. It might sound dystopian, but she had a different take. “The goal here isn’t headcount reduction,” she said. “This frees up humans to focus on more complex and strategic work which can move the business forward.” At the other end of the spectrum, sectors requiring complex problem-solving and interpersonal empathy, such as healthcare and education, stand to gain as AI augments diagnostic accuracy and personalises learning, O’Reilly argues. “The long-term strategy should be to proactively reinvest AI productivity gains back into the workforce,” she said. “By putting an intentional focus on digital upskilling alongside job redesign, we can ensure AI augments human potential rather than replacing it.” No short cuts For Chris Fong, founder of the consultancy AI Governance Operational, Singapore’s budget measures provide a promising framework awaiting the hard work of execution. The broad strokes were there, he said. “The challenge, as with many things, is in the implementation.” His prescription for workers is unambiguous. AI literacy – understanding how the tools work – is necessary but no longer sufficient. “One must become an AI native: to be proficient in the use of AI to code, to write, to learn and to develop ideas,” he said. “Be it entry-level or mid-career, it seems to be boiling down to this for staying relevant.” Laurence Liew, director of AI innovation at research organisation AI Singapore, adds a note of caution that chafes against the optimism exemplified by part-time AI certification courses. “Judgment and critical thinking cannot be developed in a weekend workshop,” he said. What Liew wants to see is “deep, immersive capacity building” delivered via programmes that hone attendees’ abilities to specify, build and evaluate AI solutions – not just gain “AI awareness”. He urged workers not to abandon their domain expertise in a rush to become technologists, though he warned of wage compression squeezing employees’ earnings – even if their jobs technically survive – as AI raised productivity levels. Small advantage Singapore is well positioned to weather the AI transition, according to observers. It was among the first countries in the world to adopt a national AI strategy, doing so in 2019, a head start that experts say is not to be underestimated. “At the framework level, Singapore is world class,” Sasson said. “Silicon Valley builds the tools, Singapore builds the system to deploy them at a national scale. No other government comes close to that level of support for pushing people forward.” On Tuesday, Manpower Minister Tan See Leng announced a new visa track launching in 2027 for “pinnacle talent” in AI and associated emerging technologies like quantum computing – a signal that Singapore intends to attract the world’s best minds even as it develops its own. Fong was emphatic on one point: the city state need not waste time building frontier AI models from scratch. “Most value comes from applying AI, not building the foundational models,” he said. “Countries that fail to adopt AI at scale risk falling behind more than those that lack domestic frontier models.” The smarter play, he argues, is to build and fine-tune smaller, efficient models that are linguistically and culturally relevant and to run them well. Singapore’s biggest structural advantage is, paradoxically, its small size, according to Liew. Large nations often face enormous coordination challenges corralling millions of people and competing interests. “In contrast, Singapore can operate as a unified system – aligning education, workforce development, industry adoption and governance in a coherent and efficient way,” he said. The city state is able to move faster, course-correct more quickly and implement policies with a precision that larger economies struggle to match. Back at his desk, Joshua Tan is still writing code: some of it his own, some of it the machine’s, all of it reviewed by someone with more experience than either of them. He does not know what his career path will look like over the next 10 years. Few people his age do. But he has arrived at something close to a working relationship with the uncertainty. “It still hallucinates,” he said of AI. “You still need a human in the loop who can be held accountable and sign off.” For now, that human is someone like him.

Singapore’s workers brace for AI disruption: ‘it’s inevitable’

.css-1c6uqr6{color:inherit;font-weight:inherit;font-size:inherit;font-family:inherit;line-height:inherit;overflow-wrap:break-word;}The city state wants 100,000 workers to become ‘AI bilingual’. But with entry-level roles already vanishing, time may not be on its side

The city state wants 100,000 workers to become ‘AI bilingual’. But with entry-level roles already vanishing, time may not be on its side