Disengagement at work may endanger Singapore’s AI push

Editor’s note: This is the second in a two-part opinion on the topic of disengagement at work. Read the other here.

Singapore is staking its next chapter on artificial intelligence, with serious investment and big ambitions because AI is the primary lever available to grow an economy with an ageing and potentially shrinking workforce. But there is a problem at the heart of this technologically driven strategy that is rarely talked about: that of growing disengagement at work among the humans.

The AI age demands curiosity, adaptability, and a willingness to reinvent oneself and how work gets done. To succeed, people must shift from being the best executors of routine tasks to becoming stewards of technology, leaning on the uniquely human traits AI cannot replicate: discernment, creativity, and an industry savviness that only comes from feeling invested in the work.

And yet the 2026 Gallup State of the Global Workplace report records just 14 per cent of Singapore’s employees as genuinely engaged – well below the South-east Asian average (25 per cent) and the global average (20 per cent). These numbers have remained stubbornly low for three consecutive years.

Disengagement is easy to miss. Disengaged workers do everything that most professionals do. They attend meetings, hit deadlines, and satisfy performance reviews.

What they do not do is extend themselves beyond the formal minimum. They do not flag systemic problems outside their remit, mentor junior colleagues without a mandate, or volunteer for ambiguous initiatives. When it comes to enterprise AI adoption, they use tools when required by their employers. But they rarely invest in experimenting, exploring how workflows can be redesigned, or sharing with teammates how to level up with AI – not without structural incentives or deliberate prodding.

Gallup’s US data captures the consequence: 65 per cent of workers in AI-implementing organisations say the technology has improved their personal productivity, but only 12 per cent strongly agree it has transformed how their organisation works. AI deployed onto a disengaged workforce produces pilots, not transformation – because enterprise-level integration sits in the hands of those same disengaged employees at every level.

What is lost is hard to quantify in dollars or cents. But Gallup estimates low engagement costs the global economy US$10 trillion (S$12.9 trillion) annually, or roughly 9 per cent of gross domestic product. Singapore too pays an invisible productivity tax every quarter – the process improvement left unproposed, the client insight never surfaced, the collaboration abandoned because the effort does not feel worth it. In an economy where every scarce unit of manpower must pull its weight, this tax rarely shows up on a balance sheet but does show up in output.

To some extent, this apathy is predictable. Remuneration and promotion frameworks in most large firms do not meaningfully reward initiative or risk-taking; they select against it. So why would professionals who run the same cost-benefit calculation their employers do reach a different conclusion?

Fixing AI-era disengagement requires changing the underlying payoff structure – demonstrating that the person who improves the workflow becomes more valuable after the improvement, not more replaceable.

That is not a common response, judging by the number of companies announcing AI-attributed layoffs without first engaging honestly with current employees about how their roles can evolve and their value be retained through upskilling.

On a more consequential level, disengagement at work is ultimately a manager-subordinate problem. Gallup’s long-running meta-analyses find that managers account for as much as 70 per cent of the variance in team engagement. But before we default to the familiar conclusion that Singapore needs better manager training, it is worth asking whether managers themselves are still in the game.

Globally, they are not. Manager engagement has dropped nine percentage points since 2022, with the steepest fall between 2024 and 2025 – from 27 per cent to 22 per cent. Managers used to enjoy an “engagement premium” over the people they led. That premium is gone.

Disengaged managers do not just underperform, they reproduce disengagement downward through their tone, their lack of follow-up, and the feedback loops they let decay. We cannot expect managers to champion a transformation they are not motivated to lead themselves.

The AI data amplifies the point. Within US organisations investing in AI, employees whose managers actively support the team’s AI use are nearly nine times more likely to say AI has transformed how work gets done – but fewer than a third report that kind of managerial backing. Whatever the Singapore-specific equivalent proves to be, the directional argument holds: AI adoption follows manager engagement, and manager engagement is in retreat.

Singapore’s problem of management quality is partly a selection problem masquerading as a training one. We promote for individual performance and technical excellence, then try to coach those promotees into people managers with strong relational skills – inside structures that still reward the original single-contributor behaviour. Then when managers fail, we diagnose it as a training gap. We are running in circles.

The harder question is not how to develop better managers, but whether Singapore organisations have designed manager roles that capable people can actually succeed in. Widening spans of control, absent upward feedback, and rewards decoupled from team outcomes are governance decisions, not training gaps. No coaching programme will repair a role that is structurally set up to fail.

I cannot emphasise enough how important it is for organisations to address the problem of low levels of engagement at work, especially after a period of consistent decline worldwide. Gallup has recorded two consecutive years of global engagement decline. No region of the world improved engagement in 2025. And Singapore has been flat for three years running.

Organisations that close this gap now will build the psychological safety, manager quality, and culture of workforce investment that successful AI adoption actually requires. Those that defer will run better technology on the same disengaged human substrate and wonder why the returns are modest.

We built this city-state on structure, discipline, and top-down strategy. What comes next requires something harder to mandate: people who are genuinely invested in the outcomes of their work. The human qualities AI cannot replicate – judgment, relational intelligence, creative risk-taking, moral courage – are becoming the only durable source of competitive advantage.

Right now, disengagement is driving those qualities underground. And in a manpower-lean economy with no room for the productivity it is leaving on the table, that is a challenge Singapore can no longer afford to defer.