AI is not killing Singapore’s offices. It’s giving some a premium

Citi Commercial Pte Ltd

IF ARTIFICIAL intelligence does more of the work, companies need fewer people, and fewer people need less office space, right?

It is a logical argument and a fairly linear expectation in the cyclical office market. However, there is a caveat to this – capital is not behaving as though it were true.

Often, we’re told to follow the money. This is what we find when we do so: In the first half of 2026, US$92.5 billion of commercial real estate investment flowed into the Asia-Pacific, a 35 per cent increase year on year. Offices commanded the bulk of the capital.

Underscoring the strength of the office market is its leasing performance in a bellwether market such as Singapore.

Net absorption of Grade A offices in the Central Business District reached about 500,000 square feet (sq ft) over the first three quarters of 2026, up from roughly 400,000 sq ft over the same period in 2025.

That means companies occupied more square footage of such offices across the one-year period.

Furthermore, Grade A vacancy stood at just 5.9 per cent in the third quarter and is on track to tighten towards 5 per cent by the end of 2027, with rental growth holding at around 4 per cent a year.

Measured against every visible metric, this is not a market pricing in obsolescence. Somewhat strangely, AI is not affecting either office leasing or real-estate capital investment in Singapore negatively.

In fact, the opposite is occurring. AI is raising the value of the right space and quietly stranding the rest.

AI-human collaboration

I believe two mechanisms are at work, with one more obvious than the other.

The clearer one is that the AI industry occupies buildings. Tracking indicates that AI companies’ office footprint in Singapore has grown to roughly 500,000 sq ft in 2026, from about 300,000 sq ft in 2025.

That’s modest in absolute terms, but a growth trajectory that reflects the city-state’s strengthening position as a regional hub for AI deployment.

The bigger story is what AI does to the work that stays human. Our analysis shows that 61 per cent of Singapore organisations expect AI to enhance human roles rather than replace them.

When routine output gets automated, what remains is judgment, negotiation, client work and complex problem-solving – the kind of tasks that are hardest to do well from a spare bedroom.

This is where the office stops being a factory for individual output and starts being the location where the high-value thinking and interaction happen.

Real-estate leaders in Singapore would agree, as many have already repriced their mandate accordingly.

When we asked which measures beyond cost and utilisation now define the value of their role’s function, 46 per cent in the Republic named employee productivity and performance – the top answer – ahead of 43 per cent across the Asia-Pacific and 39 per cent globally.

Energy efficiency and rapid reconfiguration followed. Interestingly, cost containment no longer tops the list. Put simply, a workforce engaging in higher-value work is worth more per square foot than a workforce delivering lower-value work in cheaper space.

What investors should underwrite

Here is where the opportunity sits. Four in five Singapore organisations recognise the need for portfolio action in response to AI-driven transformation.

Yet only 11 per cent are optimising for the effect of AI in their real estate operations; just 28 per cent have active programmes to redesign space for human-AI collaboration, the least advanced initiative we measure.

That demand will still materialise though, translating into leasing decisions over the next two to three years, just as major commercial redevelopments in Singapore come online in 2028.

Reading AI uncertainty as a sign of delayed demand overlooks the fact that businesses may still be preparing for future space needs.

The numbers tell a compelling story tilted in the office’s favour. Around two-thirds of businesses desire AI-driven building optimisation over basic building management; 62 per cent prioritise building quality and amenities over prime location; and 70 per cent lean towards long-term leases rather than short-term flexibility.

Deloitte’s relocation to Orchard Central is one such example. An older retail building, located in Singapore’s shopping district, with upper floors repurposed for office use is not an obvious choice on paper, but the space delivers modern functionality.

Meanwhile, prime addresses that lack infrastructure capability are losing their pricing power. Tenants have no desire to subsidise underperforming buildings.

This is especially so in Singapore, where 42 per cent of organisations surveyed identified rental increases as the steepest occupancy cost pressure, compared with 28 per cent across the Asia-Pacific.

For investors, the practical move is to audit portfolios against capability rather than classification, and to deploy retrofit capital now. The 2028 pipeline will set the benchmark for what “AI-ready” means while everything else gets repriced against it.

The constraint is internal

Our frequent interactions with occupiers in Singapore reaffirm that the binding constraint to real estate transformation is rarely budget or technology, but rather organisational friction.

More than a third cited organisational silos as the top limit on real estate value creation, 12 percentage points above the regional average, with AI and technology skills gaps at 32 per cent.

The capability exists, but the ability to mobilise it across business units does not.

This is fixable through cross-functional governance, closer alignment between C-suite strategy and real estate decisions, and shared data platforms.

It is an urgent task: two-thirds of Singapore organisations say they would prioritise long-term transformation over short-term cost minimisation. The leases being signed in 2026 and 2027 will determine whether they can.

No business wants to discover, three years into a 10-year lease, that it is paying for space its people have outgrown. AI is not ending the office as much as creating two tiers: those still worth occupying and those that are not.

The writer is country CEO, Singapore and South-east Asia, JLL


More News