Articles

Introduction

Malaysia’s approach to AI has been largely framed through the lens of economic transformation, digital innovation and public sector modernisation. For example, national development planning frameworks position AI as a tool for economic growth, public service delivery and digital governance, rather than as a core component of defence strategy.

However, recent developments such as the Defence White Paper (DWP) mid-term review, discussions surrounding cyber and electromagnetic capabilities, and the Putrajaya Forum point to a parallel yet less examined trajectory: the gradual expansion of AI into the country’s security and defence architecture.

This expansion is not occurring through a singular military AI strategy—Malaysia’s military AI trajectory is emerging through two interrelated dynamics.

First, the country’s broader AI ecosystem remains largely civilian-led, shaping access to talent, infrastructure and governance models that the defence sector increasingly relies upon.

Second, AI is beginning to enter defence institutions less as a tool for autonomous warfare and more as a potential framework for information synthesis, surveillance and decision-making.

Together, these dynamics are creating new pressures on defence planning, institutional coordination and governance in ways that remain underexamined in Malaysian public discourse.

Early Steps

Malaysia’s defence planning has long recognised the importance of the cyber and electromagnetic domains. The 2019 DWP identified these areas as increasingly critical to national security and outlined the need for the Malaysian Armed Forces (MAF) to develop capabilities to address emerging threats.

This was followed by the formal launch of the Defence Cyber and Electromagnetic Division (BSEP) in 2020, building on earlier communications and electronic warfare structures dating back to 1993.

However, Malaysia’s cyber and AI-related defence developments remain institutionally unsettled. References to cyber and electromagnetic structures across public and institutional discourse suggest an ongoing effort to align legacy structures with emerging operational requirements.

This trajectory has also been shaped by developments beyond Malaysia itself. References to contemporary conflicts such as those in Gaza and Ukraine have reinforced the operational importance for advanced technologies, including AI-enabled systems, combat drones and electronic warfare capabilities.

Simultaneously, public messaging that future conflicts may begin in the digital domain reflects a broader shift in how threats are understood, particularly as cyber, informational and operational risks become increasingly interconnected.

Malaysia’s participation in international initiatives such as the Responsible AI in the Military Domain (REAIM) process similarly suggests growing awareness of emerging governance questions surrounding military AI.

Despite this growing emphasis, the operational mandate with the MAF and the Ministry of Defence – as well as the nomenclature used to describe these capabilities – remains unresolved. This ambiguity reflects a broader challenge facing Malaysia’s defence modernisation: integrating emerging technologies into institutional structures that are still adapting to rapidly evolving security demands.

These tensions are particularly visible in maritime security, where overlapping responsibilities across the Malaysian Maritime Enforcement Agency (MMEA), Royal Malaysian Police and Royal Malaysian Navy (RMN) complicate the integration of AI-enabled intelligence and situational awareness systems.

While AI may improve information fusion and maritime domain awareness, implementation ultimately depends on coordination across civilian and military agencies.

Asymmetry

Unlike more centralised defence innovation ecosystems, Malaysia’s cyber and AI capabilities are distributedacross multiple institutions. This creates a fragmented baseline for military integration and increases coordination demands across institutions.

As a result, military adoption is occurring within an ecosystem where key capabilities, expertise and governance structures are often shaped outside the defence domain itself. This reflects a structural reality in which many foundations of AI capability development sit beyond the MAF’s immediate institutional remit.

These constraints have increased reliance on external engagements through defence diplomacy initiatives, such as training exercises, workshops and exchanges. These increasingly draw on external expertise, modern-day conflicts and evolving threat environments to support capability development.

Collaborations with Turkish defence firms exemplify efforts to accelerate technological development and remain competitive within a rapidly evolving regional security environment. However, such partnerships also introduce questions surrounding interoperability, self-reliance and governance.

These considerations intersect with Malaysia’s broader digital ambitions in important ways.

Plans to establish a sovereign AI cloud, while framed primarily around economic competitiveness, data security and strategic autonomy, also carry implications for defence data integration and technological dependence.

The integration of AI into Malaysia’s defence domain therefore extends beyond simple technological adoption and increasingly depends on how still-developing capabilities are aligned with fragmented institutional structures and competing national priorities.

However, this institutional configuration also introduces coordination challenges. While such a structure may function effectively for constabulary and regulatory roles, the integration of AI-enabled intelligence systems raises questions about how data collected and processed by civilian agencies interfaces with military command structures during security contingencies.

This institutional configuration also creates coordination challenges. AI-enabled intelligence systems often rely on data collected and processed across civilian agencies, yet translating such information into military decision-making during security incidents remains institutionally complex.

While these systems may function effectively within civilian enforcement or regulatory contexts, uncertainty surrounding jurisdiction, coordination, and operational authority can complicate responses when threats span multiple domains.

Decision Architecture

Public debate on military AI often centres on automation and autonomous action in combat. In Malaysia’s case, however, the more immediate transformation lies in AI as decision architecture.

Rather than executing military actions directly, AI systems are more likely to integrate data, prioritise threats and surface signals across complex operational environments.

Malaysia’s security architecture already places significant emphasis on surveillance, situational awareness and coordination across land, maritime, air, and cyber domains. In this context, AI is likely to shape how information is shared, escalated and prioritised across institutional boundaries.

Over time, this may influence how threats are interpreted and how responses are calibrated across defence and security institutions.

Understanding military AI in this way reframes the issue from one of automation to one of interpretation. The central challenge is not whether machines will act independently but how machine-mediated representations of reality shape judgement and escalation within increasingly data-driven security environments.

Operational Value vs Technological Momentum

AI offers clear operational benefits for defence institutions, particularly in enhancing situational awareness, accelerating response times and supporting decision-making across complex environments. However, translating technological ambition into operational capability remains uneven.

The challenge lies less in identifying use cases than in aligning technological momentum with institutional capacity.

In Malaysia, these constraints are particularly pronounced given the limited pool of specialised cyber and AI personnel within the defence sector and strong competition from the civilian digital economy. As a result, practical implementation is likely to remain incremental rather than transformative in the near term.

One area where these tensions become visible is surveillance and intelligence integration. AI-enabled systems may improve monitoring and information fusion across domains, but operational value ultimately depends on whether intelligence generated across institutions can be translated into decision-making processes during security contingencies.

This creates a broader challenge of coordination where technological capability may advance faster than institutional adaptation.

Within Southeast Asia, Malaysia’s trajectory positions it as an adaptive rather than leading actor in military AI development.

Why the Gap Matters

As AI becomes more embedded within defence institutions, questions surrounding governance, oversight and accountability become increasingly difficult to address through existing frameworks. In Malaysia, policy discussions continue to focus largely on civilian applications, while governance of defence-related AI remains comparatively underdeveloped.

This matters because AI systems increasingly shape how institutions process information. Decision-support systems influence how information is filtered, how priorities are established and how responses are calibrated.

In multi-agency environments, algorithmic outputs may begin to substitute for institutional consensus, with certain signals receiving heightened attention while others are quietly deprioritised.

A further challenge lies in testing and oversight. Without robust evaluation and accountability mechanisms, distortions in how AI systems prioritise information may accumulate over time. This is especially important within security environments where operational decisions increasingly rely on data integration and predictive analysis.

Conclusion

The challenge for Malaysia is therefore not simply technological adoption but aligning emerging capabilities with institutional structures that remain fragmented, capacity-constrained and only partially adapted to the demands of AI-enabled security.

Military AI integration is unfolding through the interaction between civilian-led technological development and defence-specific institutional realities, requiring coordination across mandates, agencies and evolving operational priorities.

More broadly, Malaysia offers an important case for understanding how emerging economies are integrating AI into defence institutions under conditions of evolving mandates and competing national priorities. Without greater clarity, AI risks being embedded across defence institutions and civil-military interfaces in ways that are difficult to coordinate, govern, or fully understand.