Malaysia’s AI winners will be determined by data activation

By David Irecki, Chief Technology Officer for Asia Pacific and Japan at Boomi

Malaysia’s artificial intelligence (AI) conversation is maturing, but the practical challenge is still the same: how do organisations turn AI investment into measurable business value?

The policy backdrop shows that AI is now a national priority. Through the National AI Office and the National AI Action Plan 2026–2030, Malaysia is building stronger foundations for governance, capability, and trust. But the real test will be inside organisations, where data is still fragmented across systems, applications, and teams.

Boomi’s “AI Ambition Meets Data Reality: APAC Technology Priorities and Challenges 2026” report found that 92% of APAC organizations are already consolidating across data, process integration, application programming interface (API) management, and automation, while 94% say data integration, access, and governance are key priorities. The direction is clear, but the real challenge is turning that intent into execution at scale.

This is why the next phase of AI success in Malaysia will not be determined by model choice alone. It will depend on whether organisations can connect, govern, and operationalise the data that AI relies on.

Adoption is not value

The same study highlights the scale of the opportunity. Malaysia is ahead of many regional peers in AI adoption maturity, with 86% of organisations reporting AI initiatives in place and 81% saying they have dedicated AI budgets.

This is a strong signal of intent. The tougher question is how many of those initiatives are actually producing measurable business outcomes.

And this is where many AI programmes run into the same problem. Much of the discussion still focuses on models, tools, and interfaces, but in practice the constraint is usually data. If the information feeding AI is incomplete, inconsistent, or trapped in disconnected systems, the output will reflect those limitations.

Customer records, transaction data, service histories, billing platforms, and marketing systems often contain different versions of the same business reality. In a pilot, teams can often work around that manually. At scale, they cannot. The inconsistency gets embedded into recommendations, workflows, and decisions.

A simple example illustrates this point. One system may define an active customer by recent purchase activity. Another may define the same customer by an open support case. A third may use campaign engagement.

Each definition may be valid on its own, but an AI agent cannot reliably act on conflicting signals without a unified and trusted data view. The result can be poor recommendations, duplicated outreach, or missed service issues.

Why data activation matters

This is where data activation becomes essential. Data activation is not just about storing information or moving it between systems, but about making that data usable across the business in real time, with the right context, governance, and connectivity so decisions can be made with confidence.

When this foundation is in place, AI becomes more useful because it is working from information that is accurate, accessible, and business-ready. It is no longer a standalone layer sitting on top of fragmented processes. It becomes part of how the enterprise runs, from customer operations to sales, finance, and service.

The organisations that are getting results from AI are not necessarily the ones deploying the most tools. They are the ones doing the harder work underneath: reducing fragmentation, improving data quality, and connecting the systems that determine how work gets done. This is the difference between experimentation and execution.

This is also why the gap between ambition and outcome is still visible in Malaysia. Many organisations are enthusiastic about AI, but fewer have integrated it deeply into core operations. The issue is not a lack of interest, but rather operational readiness.

Sectors under pressure

This challenge is especially relevant in sectors that matter to Malaysia’s competitiveness, including financial services, telecommunications, logistics, and digital infrastructure. These industries depend on speed, trust, and accuracy, but they also tend to have complex legacy environments that make integration and governance harder, not easier.

In such environments, AI success will not come from adding another layer of technology on top of old processes. It will come from designing a data foundation that can connect applications, synchronise information in real time, and support governed decision-making across the enterprise.

This is where the real work begins. If organisations want AI to scale, they need to treat data readiness as a business priority, not an IT afterthought. Governance, integration, and trust are not constraints on AI adoption; they are what make adoption commercially viable.

Malaysia is well positioned to benefit from this shift. The country has policy momentum, active enterprise investment, and a growing recognition that AI is as much an operational challenge as a strategic one.

The companies that will outperform are not necessarily those that move first, but those that build the strongest foundations and use data in a way that is consistent, governed, and repeatable.

This is the practical lesson for business leaders. AI value does not start with the model, but with the data behind it. If organisations want AI to deliver meaningful returns, they need to make that data accessible, trustworthy, and ready to support decisions at scale. This is what data activation is really about.

In Malaysia’s next chapter of AI adoption, the winners will not simply be the organisations that experiment the fastest, but rather those that operationalise the speediest. They will be the ones that turn data into decisions, decisions into action, and action into measurable outcomes.

This is where the competitive advantage will come from. And in Malaysia, as across the region, the organisations that activate their data most effectively will be best placed to capture value from AI.