When Singapore’s Non-Constituency Member of Parliament Eileen Chong stood up during the Budget 2026 debate on 24 February, she said something that stopped me mid-scroll.
She wasn’t talking about export controls, compute clusters, or sovereign AI infrastructure. She was asking a far more human question: What is the point of AI productivity if we’re just more exhausted?
Drawing on international research, Chong cautioned that “productivity gains do not automatically become human gains.” Without deliberate policy intervention, she argued, AI tools in the workplace could make jobs more intense, not less. People would be expected to do more, faster, with the same number of hours in the day. The efficiency would flow to organisations. The cost would be borne by people.
She urged Singapore’s government to think not just about how AI can make Singaporeans more productive, but about how it can give them their time back, to be present for families, to build community, to become themselves.
It was a remarkable speech. And as someone who works at the intersection of AI strategy and Australian industry, I think we need to have the same conversation here. But with one critical difference.
Singapore’s Problem and Ours: Same Risk, Different Geography
Singapore’s concern about AI and isolation makes intuitive sense when you understand its context. It is one of the most densely populated nations on earth. Most Singaporeans live in high-rise apartments. You pass neighbours in lifts. You share hawker centres. Physical proximity is built into daily life.
And yet, even there, a 2024 Institute of Policy Studies poll found that Singaporeans aged 21 to 34 report the highest levels of loneliness and social isolation of any age group. Even in a city designed for density, connection is not automatic.
Now consider Australia.
We live in houses, not apartments. Our suburbs sprawl for tens of kilometres. We drive, not walk. We work from home, and increasingly, full-time. In many Australian towns and outer suburbs, you can go several days without a meaningful face-to-face interaction. Not because of anything unusual, but because the built environment and our work patterns make isolation the path of least resistance.
If AI adoption without guardrails risks intensifying workloads and eroding community time in Singapore, the same dynamic in Australia risks something deeper: embedding loneliness into our economy as a structural feature.
We are already isolated. AI, deployed purely as a productivity tool, could seal that isolation in place.
What This Means for Australian Sectors
This isn’t abstract. The sectors I work with every day are all navigating AI adoption right now. Here is what liveability-first AI policy could look like in practice across three of them.
Defence
Australia’s defence sector is undergoing a period of significant transformation. Autonomous systems, AI-assisted intelligence, and accelerated decision-making are reshaping how the ADF operates and how defence industry partners support it. This is necessary and largely positive.
But the workforce behind these systems is made up of people, many of them in high-pressure, high-consequence roles with significant mental load. If AI adoption in defence means higher operational tempo, more outputs expected from the same personnel, and fewer moments for mentorship, knowledge transfer, and human leadership, we are building capability at the cost of the people who carry it.
Liveability-first AI policy in defence looks like using AI to reduce administrative burden on personnel so their human energy goes toward judgment, relationships, and development rather than paperwork. It looks like measuring AI success not just in capability outputs, but in retention, wellbeing, and the quality of leadership being developed for the next generation.
Hospitality
Hospitality is one of Australia’s most significant employers, and one of its most chronically overworked sectors. Rosters change last-minute. Split shifts eat into family time. Burnout is endemic.
AI offers genuine promise here: smarter scheduling, demand forecasting, and automating repetitive back-of-house tasks. But deployed purely for margin improvement, those gains accrue to operators, not workers. Staff end up with fewer colleagues, higher expectations, and the same unpredictable hours.
Liveability-first AI in hospitality looks like scheduling tools that optimise for worker consistency and predictability, not just occupancy. It looks like using AI-generated efficiency to restore hours to workers rather than extract more from fewer people. It looks like an industry that competes on being a good place to work, not just a cheap place to operate.
Local Councils
Local government is the layer of the state closest to community life. Councils run libraries, parks, aged care services, community events, and planning processes. Their staff are often the connective tissue between government and the people it serves.
And council staff are drowning in administrative load: compliance requirements, reporting cycles, grant acquittals, planning documentation. Much of this can be substantially automated using AI that is already available.
The opportunity here is profound. If councils use AI to free up staff time and then deliberately redeploy that time into community engagement, outreach, and service delivery, AI becomes a tool for social infrastructure. It does not just make councils more efficient. It makes them more present in the communities they serve. That is exactly what Eileen Chong was asking for: productivity gains that flow to people, not just organisations.
What Australia’s Government Should Consider
Eileen Chong was speaking to Singapore’s parliament. But her core argument translates directly into a policy challenge for Australian governments at every level.
First, we need a national conversation about what AI productivity gains are for. Right now, the dominant framing is economic competitiveness. That matters. But we also need to ask: will AI adoption leave Australians with more time for family, community, and civic life, or less?
Second, planning and remote work policy needs to catch up with AI reality. If AI enables more people to work from home permanently, and we do not invest in the community infrastructure that makes suburban and regional life liveable, we will have solved a productivity problem while deepening a social one.
Third, sector-specific AI adoption frameworks should include liveability metrics alongside efficiency metrics. In defence, that might mean retention and leadership quality. In hospitality, roster predictability and worker wellbeing. In local government, community engagement hours per staff member. These are not soft measures. They are indicators of whether AI is working for Australians or simply on them.
The Bigger Picture
Eileen Chong’s speech was delivered in Singapore, to a Singaporean audience, about Singaporean challenges. But the instinct behind it, that AI strategy must be grounded in human outcomes and not just economic ones, is universal.
Australia has the opportunity to lead on this. We have sectors genuinely grappling with AI adoption. We have a government, at various levels, trying to figure out what responsible AI looks like in practice. And we have a geography and a way of life that makes the stakes of getting this wrong particularly high.
The question is not whether AI will change how Australians work. It will. The question is whether we are designing that change deliberately, with liveability at the centre, or simply letting efficiency logic run its course and hoping the human costs sort themselves out.
Singapore is already asking the right question. It is time Australia did too.
G is the founder of Forge and Guild, an AI strategy consultancy working with Australian organisations in defence, local government, hospitality, and beyond. Forge and Guild focuses on strategic thinking, not just implementation, helping leaders understand what AI means for their organisations, their people, and their communities.
