The following commentary is contributed by Boomi’s Chief Technology Officer for Asia Pacific and Japan David Irecki
Conversations about artificial intelligence (AI) in the financial sector have changed noticeably over the past two years. In that time, the conversation has transitioned from experimentation — with ideas being tested and pilots being run in sandboxes — to business practicality.
In many ways, the excitement around AI has settled into something more realistic. That is not a bad thing, especially as Malaysia tries to lay claim to becoming a key player in Southeast Asia’s (SEA) AI landscape. In fact, it is exactly when things start getting interesting. Because the real challenge is no longer proving that AI works, but making it work at scale.
Malaysia’s Growing AI Momentum
According to the e-Conomy SEA 2025 report by Google, Temasek and Bain & Company, Malaysia captured 32% of the region’s total AI funding between the second half of 2024 (2H24) and 1H25 — or roughly US$759 million.
More than just enthusiasm for AI tools, this reflects a growing focus on specialised AI systems that can perform tasks within specific industries. In financial services, this could mean AI systems that help detect fraud earlier. It could involve tools that support better credit risk decisions. It may also include intelligent systems that anticipate what customers may need next.
But building these capabilities is rarely straightforward. Banks operate in complex environments. Their systems often span decades of technology, multiple vendors and large volumes of sensitive data. Introducing AI into that mix requires careful planning.
Getting Governance Right
Financial institutions operate in one of the most tightly regulated sectors in the economy. Mistakes are costly. Errors can affect customers, compliance obligations and market trust.
For these reasons, organisations must leverage the National AI Governance and Ethics Guidelines, which offers a framework to ensure fairness, transparency, safety and accountability that is central to scaling AI responsibly.
The goal is to ensure AI does not operate as an unexplainable “black box”. Institutions must understand how systems make decisions, monitor them closely and ensure the right safeguards are in place throughout the AI lifecycle.
Malaysia has also created the National AI Office, which sits under the Digital Ministry. Its role is to help coordinate national AI strategy while encouraging innovation and responsible adoption. For banks, this level of public sector scrutiny should incentivise action to build in AI governance before deployment.
Why Context Matters
However, AI governance is not just a bureaucratic endeavour. It helps ensure robust, contextual environments for AI success. Besides regulations, this affords banks precise customer behaviour, transaction patterns, and financial history, ensuring they can act with agility.
The challenge is not simply accessing data, but activating it. For AI to deliver real value, data must be unified, trusted and made available in real-time across the organisation.
This is where many organisations encounter problems. Data often sits in different systems. It may be incomplete or fragmented. Even when accessible, the same data point can mean very different things depending on context.
When institutions can bring that context together, activating data across systems and workflows, the benefits become clearer. Banks can identify unusual transactions more quickly, personalise services more effectively and simplify processes that customers often find frustrating.
In Malaysia, context also means recognising the country’s diversity. Financial services operate across English, Malay, Mandarin, Tamil, and many regional dialects. For AI to be useful, it needs to understand these differences and communicate clearly with customers from all backgrounds.
Turning AI into a Team Player
For many chief information officers, the biggest challenge with AI is not the technology itself. It is integration. Banks rarely operate on a single platform. Their data moves across legacy core systems, newer cloud applications, payment networks and third-party services.
If AI is detached from these systems, its value quickly diminishes. AI needs access to the full picture. This is where data activation becomes critical and why many financial institutions are investing in stronger integration capabilities. A modern Integration Platform as a Service can act as the connective layer between these systems.
Think of it as the infrastructure that allows data, applications and AI tools to communicate effectively with each other. Legacy systems remain in place, but they become part of a more connected architecture. Once systems can exchange information seamlessly, AI becomes far more useful. It can analyse data across the organisation rather than in isolated silos.
Governance Before Gains
The pressure to adopt AI quickly is understandable. Competition is increasing. Customer expectations are rising. New technologies appear almost every month. But moving fast without strong governance can create long-term problems.
Financial institutions that skip formal oversight risk introducing gaps in compliance or accountability. That can lead to operational risks and reputational damage.
Strong governance frameworks help prevent these issues. They create clear guardrails for how AI systems are designed, deployed and monitored. This means balancing innovation with responsibility. AI should move forward steadily, but always with proper oversight.
Building Malaysia’s AI-Enabled Financial Future
The period of small pilots and controlled experimentation is gradually coming to an end. The focus now is on bringing AI into the heart of the organisation and making it work across the entire enterprise.
As the industry moves closer to an era of agentic AI, more systems will soon be able to analyse situations, make decisions and carry out tasks with increasing independence.
The real competitive advantage will no longer come from AI models alone, but from how effectively organisations can connect, unify and activate their data across legacy platforms, cloud environments and partner ecosystems.
