To pursue responsible AI for growth, business leaders must address labour displacement

SINGAPORE’S latest Economic Strategy Review (ESR) is right to insist that artificial intelligence (AI) will be central to our next phase of growth. Its recommendations for workers who may be affected by AI, however, risk undermining both the adoption of AI and the well-being of workers.

The responsibility for coping with AI-driven disruption cannot sit largely on individual workers’ shoulders. Companies must play a key role, or they may discover that the very people they need to adopt AI will respond with caution, resistance or quiet disengagement.

In the few years running my now-defunct AI startup, I saw that the biggest impediment to AI adoption was not the technology but the middle managers – the very people recommending whether a company should adopt the technology. Managers who were employees feared being displaced; managers who were consultants or agents for companies feared their clients would use AI and drop them.

The most sensible course of action for any organisation serious about AI adoption is to consider its impact on their workforce.

When “lifelong learning” becomes a shield

The ESR midterm update calls for a national AI workforce strategy, more “flexible pathways” that blend work and training, and a nimbler skills system. It also emphasises career transition support – earlier retrenchment notices and tailored placement and training – so that workers can stay competitive as AI transforms jobs.

These are not bad ideas. They build on existing strengths such as SkillsFuture, Workforce Singapore and the National Trades Union Congress’ (NTUC) ecosystem of training and career conversion programmes. But the thrust of the ESR framing still leans heavily on the worker: workers must upskill, workers must reskill, workers must be “agile” enough to surf the next technological wave.

Lifelong learning then risks becoming a shield for organisations: as long as courses and credits are available, any displacement can be framed as an individual failure to keep up, not an organisational choice in how AI is deployed.

The reality is that few mid-career professionals with caregiving responsibilities and rising living costs can absorb serial transitions without strong employer-side commitments on job redesign, wage protection and meaningful progression pathways.

Displacement without responsibility is bad business

The ESR is candid that growth may no longer translate into the same volume of new jobs because AI allows organisations to raise productivity with fewer people. If companies respond to AI mainly as a cost-cutting tool – clearing out “middle-tier” roles and squeezing entry-level opportunities – they may see short-term margin gains but long-term erosion of trust and capability. Workers will see AI as a threat to be contained rather than a tool to be embraced, slowing the diffusion of AI.

The Business Times has already reported concerns that the ESR’s strong push for AI could “put the heat” on fresh graduates if not carefully managed. Young workers – and especially young women, who are disproportionately clustered in entry-level roles at higher risk of automation – may find that the “safety net” of starter jobs that build confidence and networks is eroded just as AI ramps up.

A firm that automates away its own talent pipeline today may find itself without the institutional memory and adaptable talent needed to implement AI effectively tomorrow. And it should not be surprised when it later struggles to recruit and retain AI-fluent staff.

The ethical – and strategic – case for corporate responsibility

Singapore’s experience to date suggests a different path is possible.

A recent analysis comparing China and Singapore shows that while China’s labour market used AI as a “harsh filter” that wiped out many exposed roles, Singapore’s labour market remained structurally stable even as AI exposure stayed high. The key difference was not the technology itself but how companies used it: from replacement to integration.

An ethical AI strategy for companies in Singapore should rest on three pillars:

• Treat AI as an organisational capability, not a redundancy machine. Organisations that focus on “AI integrators” – people who can embed tools into workflows and teams – create new kinds of roles instead of simply hollowing out existing ones.
• Share the gains from productivity. National actors such as NTUC have been clear that AI must translate into better wages, stronger job prospects and more sustainable careers, not just higher output per head.
• Plan for transitions upfront. Organisations that want to be responsible should not wait for a restructuring exercise to think about displaced staff; they should map at-risk roles early, co-design reskilling pathways with workers, and commit to fair notice, redeployment and, where necessary, dignified exit packages.

Ethically, this aligns with a basic social compact: workers whose data, tacit knowledge and day-to-day labour make AI deployment possible should not bear all the downside risk when algorithms get good enough to take over parts of their job.

Strategically, it is the only way to sustain workers’ willingness to experiment with, improve and ultimately trust AI tools embedded in their daily work. It also signals to prospective hires and investors that the organisation views AI not as a blunt instrument to cut headcount but as a long-term capability built together with its people.

What Singapore companies should do now

If the ESR’s recommendations are to translate into broad-based, socially legitimate AI adoption, organisations must move beyond compliance and box-ticking. Three practical steps stand out for boards and senior management:

• Publish an “AI and jobs” charter. Set out in plain language how the firm will approach automation, job redesign, redeployment and retrenchment, and how it will work with unions and government schemes.
• Tie AI investment to human capital investment. For every dollar spent on AI systems, commit a meaningful proportion to structured training time, mentorship and progression opportunities for current staff, not just hiring new “AI talent”.
• Build worker voice into AI roll-outs. Use existing tripartite mechanisms, joint consultative committees and employee resource groups to uncover concerns early and refine deployment plans, rather than treating communication as a one-way announcement.

Singapore’s compact system – with strong tripartite institutions and an explicit push for a national AI workforce strategy – gives companies a powerful enabling environment to do this. The ESR offers scaffolding; it is up to organisations to build it up such that AI augments people, rather than making them expendable.

If businesses fail to shoulder that responsibility, workers will draw the rational conclusion that every AI pilot is a prelude to the next redundancy exercise.

In that world, the promise of an “AI-empowered economy” will remain just that: a promise, not a lived reality for most workers or organisations.

The writer is professor emeritus at the Wee Kim Wee School of Communication and Information, Nanyang Technological University