PRIMALCOM ENTERPRISE

New data: 38% of Malaysian businesses are now using AI consistently — up from 27% last year. But 67% of them are stuck at the same stage: basic chatbots, off-the-shelf tools.



Only 19% have a formal strategy to scale AI across the whole business.



If you're wondering "we adopted AI, so why hasn't much changed" — don't rush to blame the AI.



In our experience, it's almost always the same root cause: the AI isn't the problem. It's the pile of messy spreadsheets, the decade-old legacy system nobody wants to touch, and the data scattered across departments that never talk to each other. Even the smartest AI can't see through a mess like that.



That's the layer nobody wants to deal with — systems integration, data modernization, connecting legacy systems to modern AI. That's exactly what we do.



Where's your AI stuck?

Drop it in the comments — let's see if it's the same problem.

⚠️ Companies are spending heavily on AI.

Far fewer are asking whether their people are ready to turn it into business results.

As Kosmo Digital reported today, citing a regional industry study, 83% of companies in Asia Pacific have had to pull back an AI system they had already launched.

And the reasons go beyond the technology itself.

Culture. Trust. Capability. Change management.

The same article also cited an estimate that around 93% of AI budgets go into technology, while only about 7% goes into culture and change management.

It also referenced research suggesting that only around 40% of AI rollouts in the region end up exceeding expectations.

The obvious conclusion is:

This is a people problem, not just a technology problem.

✅ Train the team.

✅ Build trust.

✅ Manage the change.

That is true.

But it is only half the story.

Because even when the people are ready — even when they have been trained and trust the tool — there is a second wall that training alone cannot solve:

Can the data, workflows and systems underneath the business actually support what people are now asking AI to do ❓

A well-trained team, pointed at data scattered across five systems that do not talk to each other, can still end up with a stalled AI project.

Just a better-trained one.

AI readiness is not one checklist.

It has to happen on both sides:

❓ Are your people ready to work with AI ❓

❓ Are your processes, data and systems ready for AI to work with them ❓

The mistake is treating one as a substitute for the other.

⭕ Sometimes the fix is training.

⭕ Sometimes it is workflow redesign.

⭕ Sometimes it is governance.

⭕ Sometimes it is integration.

⭕ And sometimes the data underneath needs fixing before any of the above matters.

AI adoption is not simply a tool-selection problem. It is an operating-model problem.

If your company is already investing in AI, where has the real bottleneck been so far — people, process, data or systems ❓

Happy 69th Hari Merdeka 2026!

May Malaysia continue to grow, prosper, and move forward together in unity and harmony

Malaysia doesn't have an AI ambition problem.

It has an AI readiness problem.

Malaysia's Digital Action Plan 2030 sets a clear direction: move from consuming technology to producing local digital innovation.

That is a positive signal.

But a national AI strategy does not automatically make an enterprise AI-ready.

Our view at Primalcom is simple:

Before selecting an AI model or approving another pilot, assess four layers.

🖥️ 1. Systems
Where does the required information live—ERP, CRM, spreadsheets, WhatsApp or disconnected applications?

📊 2. Data
Is the data accurate, accessible, structured, permissioned and owned by the right team?

🔄 3. Workflow
Is the process consistent, or does every department, branch or employee handle it differently?
Where do exceptions and human approvals occur?

👤 4. Operating ownership
Who will maintain the integrations, monitor the output and resolve problems after go-live?

An AI demo can work while production still fails.

If these four layers are unclear, the business may automate inconsistent data, create more manual work or become dependent on a system nobody internally can manage.

The first question should not be:
"Which AI model should we use?"

It should be:
"Which business process is ready to improve—and what must be fixed underneath before AI can operate reliably?"

Primalcom approaches AI initiatives as a business, systems, data and delivery decision—not simply as a standalone feature.

↓

Which layer creates the biggest barrier in your organisation: systems, data, workflow or ownership?