By Alvin Aloysius Goh
A national push, a firm-level hesitation
Singapore has been very clear about where it wants to go on AI. Budget 2026 set up a National AI Council chaired by the Prime Minister, a Champions of AI programme for enterprise transformation, and a merged WSG–SSG agency that will integrate support for workforce planning, job redesign and training.
The ambition is national. The scaffolding is in place.
But inside organisations, a basic question remains unresolved: Who owns the capability of the workforce to keep pace with AI?
As AI changes jobs, skills and the way work gets done, who is accountable for ensuring people can actually do the new work?
In Singapore National Employers Federation (SNEF)'s survey of 320 companies this year, 48% plan to explore, adopt or enhance AI in 2027, up from 39%. At the same time, those citing the cost of upskilling and reskilling as a challenge rose from 23% to 30%.
IDC research also found that 56% of Singapore firms cite budget constraints in reskilling, the third-highest rate among 22 markets surveyed.
Appetite for AI is rising. So is the cost of preparing people for it.
I don't think that is simply a funding problem. I think it is an ownership problem.
Two arguments, one thread
In my last two articles, I wrote about what capability looks like at the individual level.
Last week, I argued that learning should show up in what HR professionals can do differently, not simply in the courses we have attended.
Two weeks earlier, I wrote about certification. From 1 July 2028, organisations with more than 200 employees will need at least one HR professional certified at IHRP-CP level or higher, with the certified pool expected to grow from about 9,000 to around 18,000 by 2030.
I welcomed the floor, but argued that the credential opens the door. It does not walk you through it. We should measure movement, not just motion.
I also admitted that I had let my own IHRP-SP lapse because I had not made the time to submit my CPD hours. I have since been recertified as an IHRP Senior Professional with Institute for Human Resource Professionals (IHRP).
I could not keep writing about learning and standards without meeting them myself.
Those articles asked what capability looks like in an HR professional.
This one asks the same question of the organisation.
As AI changes the work, who owns the organisation's ability to keep learning, adapting and building the capabilities it will need next? Too often, the answer is nobody.
The seat nobody is sitting in
Ask CEOs and most would acknowledge that their CHROs should own the organisation's future capability. Yet the data tells a different story.
According to IBM's 2026 CHRO Study, 46% of organisations have not involved the CHRO in defining AI strategy. Only 28% of CHROs reported having a joint roadmap with IT.
In a Stanton Chase survey of 214 executives, CHROs owned AI outcomes in just 2% of organisations.
Gartner's October 2026 research also found that only 38% of CHROs and CIOs shared an understanding of how AI affects the future of work.
More concerningly, the research found that ownership of AI outcomes was becoming less clear as more executives got involved, while CIOs lacked confidence in leading end-to-end AI transformation.
Ask most CEOs who own the organisation's future capability and they will say the CHRO.
The data tells a different story.
- HR is often outside the room. IBM's 2026 CHRO Study found that 46% of organisations do not involve the CHRO when defining AI strategy. Only 28% of CHROs report having a joint roadmap with IT.
- HR almost never owns the outcome. In a Stanton Chase survey of 214 executives, the CHRO owns AI outcomes in just 2% of organisations.
- Technology leaders are not claiming it either. Gartner's October 2026 research found that only 38% of CHROs and CIOs share an understanding of how AI affects the future of work. It also found that ownership of AI outcomes is becoming less clear as more executives get involved, while CIOs lack confidence in leading end-to-end AI transformation.
Put these together and a pattern appears.
AI tools have owners. AI capability does not.
The technology sits with IT, the budget with finance, and the use cases with the business units. But whether people can actually do the new work often falls between them.
This is where the cost problem begins.
When nobody owns capability, upskilling is bought as a line item rather than built as an asset. One survey of 600 HR budget holders found that 47.1% of companies increased AI spending while cutting learning and development budgets.
Tools get funded. The people who must use them are left to catch up.
The mirror we have to look into
It would be convenient to stop there and argue that HR deserves the seat. The evidence does not allow us that comfort.
In the same IBM study, 72% of organisations make limited or no use of AI within HR itself. CHROs rated HR's AI literacy capability at 13%, AI performance measurement at 16%, and change management for AI adoption at 20%.
This is not about blame. HR has spent two decades being asked to run lean, compliant and efficient, and was rarely resourced to experiment.
But it does mean the seat will not be handed over simply because of a job title. It has to be earned through visible capability.
Last week, I borrowed Peck Kem Low's image of the cobbler whose children have no shoes: HR looks after everyone else's development and neglects its own.
On AI, that image fits uncomfortably well.
I also wrote about "the AI tool that was rolled out but never used" as a signal HR should learn to notice. Too often, that signal points back at us.
A profession that asks others to reskill cannot credibly lead this work while its own practice remains untouched.
The IBM data also shows why this matters. Where the CHRO shares responsibility for deciding which decisions remain human-led, 76% of employees feel safe questioning or overriding AI recommendations. Where HR is only advisory, that falls sharply to 43%.
Human judgement does not protect itself. Someone has to design for it.
What Singapore's data shows, and what it does not
Almost every figure on who owns AI capability comes from global or Western samples. Singapore's own data tells us a lot about adoption, role changes and cost, but very little about ownership.
What we do know points in the same direction:
- Jobs are already being redesigned, often without a designer. IDC found that 91% of Singapore organisations reported AI-related role changes or displacement. Singapore also recorded the highest displacement rate among the Asian markets surveyed, at 48%.
- AI is becoming part of ordinary roles. AI-related postings rose to 5.3% of all roles in 2025, from 3.3%, with most growth coming from non-specialist roles requiring working fluency with AI tools.
- The learning pathway at the entry level is narrowing. In the same IDC research, 75% of Singapore firms said shrinking entry-level pathways make it harder to develop future leaders, while 70% reported fewer on-the-job development opportunities for junior staff.
That last point matters.
When AI absorbs the routine tasks through which people once learnt their craft, organisations save time today and borrow judgement from tomorrow.
I call this Experience Debt.
It does not show up on a cost line, which is exactly why it needs an owner.
The national system is ready to help through workforce planning, job redesign and training support. But someone inside the organisation still has to connect it all and carry it forward.
That is the missing piece.
A bridge builder, not a bridge
The answer, in my view, is not another C-suite title.
It is a mandate: the organisation capability mandate.
The CHRO's established remit includes ensuring that the workforce can deliver the organisation's strategy. The capability mandate extends that responsibility further into the future and across functions.
It anticipates the capabilities the organisation will need three to five years from now and builds them across people, AI and ways of working.
The distinction I keep returning to is this: the leader holding this mandate is a bridge builder, not a bridge.
A bridge sits between the CEO, CHRO, product and innovation leads, and gets walked over by each of them. A bridge builder designs how they connect.
That leader helps determine where capability is built, who funds it, and how judgement is protected as routine work moves to machines.
Any senior HR leader can grow into this mandate, and in many organisations, the CHRO will be its natural owner.
But it will not be won through goodwill. As I argued in my certification piece, HR earns its seat at the table with a business case, not a bottle.
On AI, that business case is a capability plan: which jobs are changing, what skills they need, what that costs against the tools being funded, and which decisions must remain human-led.
It also takes the moral courage to say what others may not want to hear:
"We have funded the tool, but not the people who must use it."
The mandate has to be stated, resourced and measured.
Assumed ownership is how we arrived at 2%.
Where this goes, and where to start
The next few years will make the ownership question harder to avoid.
In the near term, organisations without a named owner for capability risk continuing to fund tools while upskilling stalls.
Over time, the difference will show up in performance. Firms that redesign jobs deliberately will pull ahead of those whose jobs are redesigned by default, one tool at a time.
Further ahead, Experience Debt will come due. Leaders who never did the formative work that AI absorbed will eventually reach senior roles. Organisations that protected judgement-building work will have a leadership bench. Others may have to buy one.
For HR leaders, three moves are within reach now:
- Map where AI is already reshaping roles. Start with the three to five workflows where human judgement carries the highest stakes. Redesign deliberately rather than inherit whatever each tool deployment leaves behind.
- Put a name against capability. Agree with your CEO who decides, who funds and who is accountable for AI capability outcomes. If the honest answer is nobody, that is your opening.
- Go first in your own function. Use AI visibly within HR before asking the business to. Credibility is built by demonstration, not by position.
A disclosure, since SHRI has an interest here.
We are building tools for this work, including our Workforce AI-Readiness Index. But one gap concerns me more than any product: Singapore still has very little local evidence on who owns AI capability inside our organisations.
SHRI intends to help close that gap, so our decisions can rest on our own context rather than imported averages.
There is also an opportunity in Singapore's certification push that we should not miss.
By 2030, around 18,000 certified HR professionals will sit inside our organisations. If they hold only the credential, we will have counted motion.
If a meaningful share takes up the capability mandate, Singapore will have a cadre of HR professionals able to turn national AI support into organisational capability.
That would be movement.
Last week, I argued that HR was never just a job.
AI is now testing that claim at organisational scale.
Meeting that test asks HR professionals to go higher in value creation, wider in perspective and deeper in acumen.
Organisation capability is where all three meet.
