Today, whenever you hear the word layoff, the word AI is often not far behind. Around the world, recent headlines increasingly suggest that AI is coming for people’s jobs. It is a striking narrative, but one that can oversimplify what is actually driving many workforce decisions. The reality is more nuanced. Many so-called “AI layoffs” are not simply the result of machines becoming capable of performing entire roles end to end. Today’s systems remain some way from fully replacing many of the jobs being cut,
and decisions around redundancies are often still shaped by more familiar pressures such as cost-cutting, over-hiring, restructuring and broader economic conditions. In many cases, AI has become a shorthand for changes that are still rooted in traditional business realities. In Singapore, where skills policy and productivity goals shape workforce transformation, this narrative carries heavier implications. Because when we simply blame layoffs on AI, we risk missing what’s actually changing in Singapore’s labour market, and choosing the wrong solutions. Wrong diagnosis, wrong solution I’m not denying
that AI is disrupting the workplace, but most jobs aren’t simply disappearing – they’re being redesigned, tasks are being redistributed, and it’s changing what makes employees valuable. AI isn’t replacing workers, but augmenting capabilities and redefining how work is being done. But in response to the flawed “AI layoffs” narrative, many companies assume that tools alone will deliver transformation. They end up doubling down on technology, rather than investing in helping their people learn how to use these tools effectively. They then make the mistake
of protecting old job roles and keeping ways of working the same while simply adding AI tools on top, instead of redesigning how the work should be done. By over-investing in tools and under-investing in capability-building, organisations will find shallow or uneven adoption of AI across different departments. When work is reorganised faster than companies can reskill their people, there will be a growing divide within organisations between those who can use AI to amplify their productivity, and those left doing tasks that gradually lose
value. Over time, this entrenches inequality between those with AI fluency and those without. Organisations will struggle to fully realise gains from their technology investments. This uneven capability carries a real risk in Singapore’s high-skill, high-cost economy. As past waves of technological change have shown, productivity gains do not automatically lead to better outcomes for workers. In the 1990s, when there were technological advancements like robotics, automated machine controllers and manufacturing tracking systems, the expertise sat with supervisory roles, rather than the production line. As
a result, value was concentrated at the management level, and these wages grew faster than that of production staff. Historically, increased output has often benefited a small minority rather than the workforce at large, and a similar pattern in AI capability risks creating a structural productivity ceiling. The result will not just be productivity differences within organisations, but entrenched inequalities in the labour market, where opportunities and rewards are increasingly uneven. Building broad capability So how, then, can AI capabilities be developed evenly across the
workforce? As a start, it involves rethinking how training is done. In Singapore, AI adoption is already widespread and AI fluency is being actively supported through national initiatives. Despite this, many organisations continue to invest heavily in AI tools while underinvesting in the people expected to use them – because while platforms and infrastructure are visible and measurable, workforce capability is not. It takes time to build, is harder to quantify and rarely appears clearly in financial reporting. Many companies fall into the usual patterns
with technology: Buy first, train later. But in practice, “later” often means “never properly” or sometimes not at all. Employees are expected to learn on the job. The outcome is predictable. Certain employees become highly effective at using AI, while others are left in the dark, either intimidated or uncertain about how the tool can be used in their day-to-day. This is down to a lack of access, exposure and organisational support. Even where there is training, it’s often outdated and outpaced by AI, which
is embedded into daily workflows and constantly evolving. Periodic, one-off instruction will no longer suffice. Training must shift to continuous, work-integrated learning, where employees build capability as tasks change, rather than retrospectively. This means treating capability as infrastructure, taking a “learning by doing” approach and making AI fluency a leadership expectation. For example, one global biotechnology company with a presence in Singapore made it a point for senior leaders to take the lead in learning and applying AI to their work, before expanding this practice
to all their teams. The company ensured the shift in working practices was led from the top, and all employees learnt how to use AI by completing real tasks and applying what they learnt to real business problems in their jobs. To go one step further, giving employees real-time updates on their AI skill levels, usage and progress will encourage them to keep experimenting with AI tools as part of normal work. Rather than focusing on whether AI replaces jobs, the real issue to focus
on is how it is reshaping jobs, rewarding those who can use it and disadvantaging those who cannot. The real divide is no longer between employment and unemployment, but between AI-enabled and non-AI-enabled workers. And that divide will continue to widen unless organisations can shift how they think about investment, building capability as deliberately as they deploy technology. Misdiagnosing that shift by blaming job losses on the technology alone could redirect attention and solutions to the wrong problem.
Singapore, AI layoffs, workforce transformation, skills policy, productivity goals, capability building, inequality, labour market