Malaysia solved the easy part of AI.
The hard part is still sitting inside your company.

“Malaysia now has an AI plan.” That’s the headline on a Free Malaysia Today opinion piece published today, 3 July, by columnist Murugason R Thangaratnam. His argument is worth sitting with: Malaysia Digital 2030 is a sound plan on paper, five clear pillars, real targets, real ambition. But as he puts it, “the real question is not whether we can write a good digital plan; it is whether the country can turn that plan into better services.”

He’s right, and he’s specific about where it gets hard. Too many digital initiatives, he warns, “remain trapped in silos, never scaling into a coherent system that citizens can actually feel.” Trust has to be earned, not assumed. Infrastructure that looks good in national coverage numbers can still leave entire communities behind. And on talent specifically: MD2030 wants to reskill 700,000 workers and pull in 2,000 AI specialists, but that build-out has to reach past the big cities or the country ends up with a sharper digital divide, not a smaller one.

That’s the hard part at the level of a nation. Zoom into a single company, and it looks different. It isn’t “can government policy reach 700,000 workers.” It’s “does any individual leadership team actually know what its own people already know.” Microsoft answered that question a few days before FMT’s column ran, and most Malaysian leaders haven’t caught up to the answer yet.

The workforce isn’t the hard part. It’s already ahead.

The 2026 Work Trend Index, released 23 June, found that 24% of Malaysian workers already qualify as “Frontier Professionals”: advanced AI users who’ve rebuilt how they work around it. The global average is 16%. Malaysian employees aren’t behind the world on AI. They’re beating it.

Read those together. A quarter of your workforce is operating at the global frontier. 69% of Malaysian AI users say they’re producing work they couldn’t have created a year ago. And then: barely a third can see a clear AI strategy above them, and fewer than 1 in 5 feel rewarded for changing how they work.

Microsoft has a name for this: the Transformation Paradox. Employees adopt fast. Organisational systems don’t. And the report’s most important finding is buried in the middle of it: organisational factors matter roughly twice as much as individual behaviour for whether AI actually delivers impact.

This is the hard part FMT is pointing at, at ground level. Not a national rollout problem. A visibility problem. On 24 June, Human Resources Minister R. Ramanan told Parliament that TalentCorp expects 697,000 Malaysian jobs to be highly affected by AI, digitalisation and the green economy within 3 to 5 years, close enough to FMT’s 700,000 that it’s plainly the same cohort. The government has clocked the scale of the problem. What it can’t do for you is tell you where your own people stand today.

The hard part was never getting people to use AI. It’s building a company that can see who’s already good at it.

The fix is structure, not another town hall

If organisational factors matter twice as much as individual behaviour, then the fix lives at the organisational level. Practically, that means 4 things.

Measure where people actually are. Not who claims to use ChatGPT. Actual assessed capability, person by person, so you know where your frontier is and where your gaps are.

Train to application, not videos. Your frontier workers didn’t get there by watching content. They got there by applying AI to their real work. Training for everyone else should look the same.

Track it like you track anything that matters. Progress, engagement, capability over time. If AI skill isn’t on a dashboard somewhere, it isn’t being managed.

Reward it visibly. The 19% number is the cheapest fix on this list. People who reinvent their work and get recognised for it teach everyone around them. People who don’t, leave.

This is where an AI adoption strategy for Malaysian corporations stops being a slide and starts being a system.

From individual enthusiasm to business capability

FMT is right that Malaysia’s AI plan will be judged on execution, not announcements. Execution, inside a company, starts with knowing what you’re working with. The Work Trend Index just handed Malaysian leaders that answer for free: your people are willing, a quarter of them are already excellent. The hard part isn’t producing more of that enthusiasm. It’s building the system that can see it, direct it, and reward it.

That system is exactly what SkillTrainer AI is built to be. AI-powered assessment agents show you where every person actually stands, not where they say they stand. Our platform trains people by applying AI in real workflows, the same way your frontier workers learned. The analytics dashboard puts capability on a screen leadership can act on, and certificates give people the recognition the data says they’re not getting.

We’re HRD Corp registered and claimable under SBL-Khas, so the levy you’re already paying can fund it. But the funding isn’t the point this time. The point is visibility.

Malaysia has a plan now. Whether your own company can see the workforce it already has is the harder question, and it’s the one nobody else can answer for you.