You are 40, earning more than you did five years ago.

Your career looks stable. Your spouse is working too. Your current home has appreciated, and the children could certainly use another bedroom.

Then the bank tells you how much you can borrow.

Suddenly, the bigger condominium you thought was slightly out of reach looks possible.

The natural reaction?

If the bank says we can afford it, why not stretch a little more?

For years, that calculation may have felt reasonable, particularly for mid-career professionals expecting their income to remain stable or continue rising.

But artificial intelligence is one more reason to examine an assumption sitting quietly behind that decision.

Not necessarily whether you will still have a job.

But whether your income five or ten years from now will follow the path you expect it to today.

That matters when the financial commitment you are considering could last 20 to 30 years.

Much of the conversation around AI and employment tends to swing between two extremes.

One says AI is going to replace huge numbers of workers.

The other says AI will simply make everyone more productive.

Singapore's experience so far appears more nuanced.

In April 2026, the Ministry of Manpower (MOM) reported that 28.5% of firms had adopted AI in some form. Among those firms, only 6.2% reported reducing headcount.

Far more were changing the work itself.

About 18.9% had redesigned job functions, while 13.9% had created new AI-related roles.

AI adoption was also particularly high in sectors containing many PMET roles:

Source: MOM, stats.mom.gov.sg

MOM's conclusion was not that AI was causing widespread unemployment. Rather, its effects were appearing more strongly in how jobs are being redesigned than in jobs disappearing altogether.

For a homeowner, that distinction is important.

Because you do not have to become unemployed for a large mortgage to become uncomfortable.

Your role could change.

Your company could restructure.

Your next position could come with a different compensation package.

You could spend several months transitioning between roles.

You might decide that staying employable requires retraining or moving into a different function.

Or a portion of your pay that once felt dependable, such as bonuses or variable compensation, could become less certain.

For a long-term borrower, the more useful scenario to test is therefore not simply whether AI eliminates the job, but whether a role change, transition or compensation reset could interrupt the income path the household had assumed.

And that is where the property question becomes interesting.

When buyers think about upgrading, the focus is usually on the property.

Can we afford the downpayment?

How much can we borrow?

What will the monthly instalment be?

Can we stretch another $200,000 for the larger unit?

But underneath all those questions sits another assumption:

that your future earning power will continue supporting today's property decision.

Suppose a couple earns a combined gross income of $20,000 a month.

Based on their current finances, they are considering taking a $1.5 million housing loan over 30 years.

At an illustrative interest rate of 3%, their monthly mortgage would be approximately $6,300.

At today's income:

$6,300 $20,000 = about 32% of gross household income.

That might feel manageable.

But imagine their combined income falls by 20% after one partner goes through a career transition.

Their mortgage has not changed.

Their property has not changed.

But now:

$6,300 $16,000 = about 39% of gross household income.

If the household temporarily relies on a single $10,000 income, the same mortgage consumes approximately 63% of gross income.

Illustrative example only. It excludes other debts, expenses, CPF usage and individual financial circumstances.

Nothing happened to the mortgage.

What changed was the income supporting it.

That is the part of affordability an approval letter cannot predict.

Singapore's Total Debt Servicing Ratio, or TDSR, requires a borrower's total monthly debt obligations to stay within 55% of gross monthly income.

It is an important safeguard against excessive borrowing.

That safeguard is more conservative than the headline 55% suggests.

For residential property loans granted by financial institutions, the TDSR calculation uses the higher of the loan's thereafter interest rate or a 4% medium-term interest-rate floor. This means a borrower may be assessed at a higher rate than the promotional mortgage rate actually being offered.

Variable income such as bonuses, commissions and allowances is also subject to a minimum 30% haircut when calculating TDSR. In other words, only part of income considered less dependable is recognised for the affordability test.

So the framework already builds in protection against higher rates and fluctuating income.

But there is still something it cannot determine.

Whether the fixed salary feeding that calculation will look the same five or ten years later.

That is the distinction.

TDSR can stress-test the financing assumptions available today.

It cannot tell you what your profession, employer or earning path will look like eight years from now.

CPF Board similarly advises buyers to think beyond the maximum loan available, factoring in job stability, future expenses and potential income changes when deciding what is genuinely affordable.

That becomes relevant when industries themselves are changing.

In June 2026, Shopee cut jobs in Singapore, with software engineers among those affected, as parent company Sea continued investing in AI.

Importantly, Shopee did not say those employees were directly replaced by AI. The company said departments may adjust staffing based on operational and business priorities following regular reviews of business operations.

So it would be inaccurate to conclude that AI simply "took their jobs".

But the episode illustrates something more relevant to a long-term borrower:

Even highly skilled, seemingly secure white-collar roles can be affected as companies reorganise around changing technology and business priorities.

That does not make those careers unsafe.

It does make it harder to assume that the way a role looks today will remain unchanged throughout a multi-decade mortgage.

This may be the more useful AI question for someone considering a larger home.

Imagine your job still exists five years from now.

That does not necessarily mean the economics of the job remain identical.

If AI allows a team to produce more with fewer manual processes, companies may reorganise responsibilities even without eliminating an entire occupation.

Some employees may become considerably more productive and valuable.

Others may find certain tasks automated or redistributed.

New roles may emerge.

Old responsibilities may disappear.

And the skills employers value most may change.

MOM's findings already point towards this process. Firms adopting AI have so far been more likely to redesign jobs or create new AI-related roles than to reduce headcount.

NTUC has raised similar concerns around adaptation.

Ahead of Budget 2026, it called for stronger transition support for PMEs as AI reshapes workplaces and employer expectations. Its Survey on Economic Sentiments 2025 found that 56% of PMEs felt they needed to upskill to remain relevant.

That does not mean those workers will lose their jobs.

Nor does MOM's evidence establish that AI is already reducing PMET salaries or long-term earning power.

What it does show is that work is changing.

For a 38-year-old considering a 25-year mortgage, that is worth accounting for.

You are not only asking whether you will remain employed.

You are asking whether your household finances have enough room to absorb the career transitions that could occur over the life of the loan.

Someone in their late 30s or 40s may actually be in their strongest financial position yet.

Your salary may be near its career high.

You may have built substantial CPF savings.

Your first property may have appreciated.

Your borrowing capacity could be significantly greater than it was ten years ago.

Which makes stretching for the bigger home especially tempting.

But this life stage can also come with heavier financial commitments.

Children.

Parents.

Insurance.

Retirement.

Education costs.

Perhaps even plans to slow down professionally later in life.

At the same time, you have fewer working years available to recover from a major financial setback than someone buying at 28.

That means your maximum borrowing capacity and your sensible borrowing capacity may not necessarily be the same number.

AI does not create that risk.

But it gives buyers another reason to think seriously about the gap between the two.

Most buyers know to consider what happens if mortgage rates rise.

In the AI era, it may also be worth asking what happens if your income changes.

Before committing to the upgrade, try running the property through several different versions of your career.

Test 1: What if household income falls by 20%?

Do not assume retrenchment.

Perhaps one partner moves into a different function after a restructuring and accepts a lower salary.

Perhaps variable compensation falls.

Perhaps a new role trades income for greater stability.

Would the property still feel comfortable?

More importantly, could you continue saving for retirement and other goals after paying the mortgage?

Test 2: What if your bonus disappears?

For some PMET households, bonuses quietly subsidise their lifestyle.

Holidays, enrichment classes, investments, renovations and annual expenses may all depend on them.

This is also why TDSR does not treat variable income in the same way as fixed salary.

Try applying the same conservatism to your own property budget.

If your mortgage only feels comfortable because you expect a strong annual bonus, ask whether that income should really support a 25-year commitment.

Test 3: What if one person needs six months to transition?

AI-related restructuring does not have to mean permanent unemployment.

The problem may simply be the gap between roles.

Could one income temporarily support:

• the mortgage;
• household expenses;
• insurance;
• children's expenses;
• support for parents; and
• other debt commitments?

If not, how much cash would you need to bridge that period?

Test 4: What if you need to retrain?

A career transition can involve more than lost salary.

You may choose to spend time acquiring new skills.

You could temporarily accept less demanding work.

You might move from a mature function into a growing one and take time to rebuild seniority.

That flexibility is easier to exercise when your household does not require every dollar of your current salary simply to maintain the home.

This is where the property decision and AI conversation connect most directly.

A cash buffer is normally described as protection against emergencies.

But for a mid-career worker navigating technological change, it can also buy time.

Time to search for the right next role.

Time to reskill.

Time to avoid accepting the first job available simply because next month's mortgage is approaching.

CPF Board recommends maintaining three to six months of expenses for emergencies and suggests retaining some Ordinary Account savings as a housing safety buffer.

For a PMET working in a rapidly changing industry, that raises an important question before upgrading:

After completing the purchase, how much flexibility will actually remain?

If almost all your cash and CPF resources are required to make the upgrade work, then the home may be affordable under today's employment conditions but less resilient if those conditions change.

This runs against how many people naturally approach property.

Once buyers discover they qualify for a certain loan quantum, that amount often becomes their new property budget.

If you can borrow $1.5 million, you start viewing homes that require something close to a $1.5 million loan.

Otherwise, it can feel like you are "wasting" your borrowing power.

But consider turning the logic around.

What if maximum borrowing capacity is a ceiling rather than a target?

Suppose a household qualifies for a $1.5 million loan but chooses a property requiring only $1.2 million.

That difference could mean:

• lower monthly repayments;
• more cash preserved after completion;
• less dependence on bonuses;
• more CPF left for retirement;
• greater ability to survive a career transition;
• more freedom to retrain or change industries;
• less pressure to immediately replace a lost salary; and
• less likelihood of needing to sell the property at an unfavourable time.

The trade-off might be a smaller unit.

A different location.

An older development.

Or simply postponing the next upgrade.

That can feel like settling.

But there is another way to view it.

You are purchasing financial flexibility alongside the property.

And in an economy where the value of different skills can change quickly, flexibility itself becomes an asset.

But that flexibility is not free.

If both careers progress as expected and the larger home appreciates, buying less could mean giving up some asset growth. If the family later needs to upgrade anyway, it could also mean another round of transaction, moving and renovation costs.

So the argument for leaving borrowing capacity unused is not that it will necessarily produce the better financial return.

It is that the two mistakes can have very different consequences.

Under-buy, and you may have to upgrade later.

Overstretch and suffer an income shock, and you may be forced to sell when the market, your Seller's Stamp Duty position or your personal circumstances give you very little negotiating room.

That is the asymmetry worth thinking about.

You may never need the additional financial buffer.

But if you do, its value could be far greater than it appeared on the day you chose the smaller mortgage.

No.

That would be an overreaction to what the current evidence actually shows.

Singapore has not experienced broad-based AI-driven job displacement.

And the wider labour market is not signalling a white-collar employment crisis either.

In March 2026, resident unemployment stood at 2.9%. There were still 1.46 job vacancies for every unemployed person, while the six-month re-entry rate among retrenched residents rose to 60.7%, with improvements seen among PMETs and degree holders.

More recent MOM statements have also continued to emphasise that firms adopting AI are, so far, more commonly redesigning jobs and creating roles than reducing headcount.

AI may make many workers more productive, create new opportunities and increase the value of people who learn how to use it effectively.

So the lesson is not:

AI is coming. Buy a cheaper home.

It is:

AI is one more reason not to assume that today's role, salary structure and career path will remain unchanged for the next two decades.

Your housing decision should be able to tolerate some uncertainty around that path.

There will still be households where stretching makes sense.

Perhaps both incomes come from different sectors.

Perhaps the mortgage remains manageable on one salary.

Perhaps substantial cash reserves remain untouched after the purchase.

Perhaps there are few other financial commitments.

Or perhaps the larger home meets a genuine long-term family need that justifies the higher cost.

The question is not whether stretching is automatically wrong.

It is whether you understand what has to continue going right for that larger mortgage to remain comfortable.

For years, career progression made one property assumption especially comfortable.

Earn more.

Upgrade.

Earn more again.

Upgrade again.

For many people, that progression may still happen.

But AI is introducing another variable into the equation.

Jobs may remain while tasks change.

Roles may survive while the skills required to perform them evolve.

Entire professions may not disappear, but individual workers could still experience periods of disruption as companies restructure around new technology.

That means a property buyer today may need to think beyond current affordability.

Your bank can tell you how much home today's salary allows you to finance.

But a 25-year mortgage will live through far more than today's salary.

In the AI era, the strongest property decision may therefore not be buying the biggest home you qualify for.

It may be buying the home that still works even when your career does not go exactly according to plan.