Malaysian Banks Struggle to Scale AI Due to Core System Limitations

A recent study indicates that only 24% of Malaysian banking institutions are successfully scaling artificial intelligence across their entire enterprise, highlighting a significant divide between strategic ambition and operational reality. While nearly every banking leader confirms AI is central to their future roadmap, the practical implementation remains fragmented across the majority of the sector.

According to the original publisher, a study conducted by Sabio World earlier this year across 14 Malaysian banking institutions revealed that 76% of respondents are restricted to running isolated AI pilots or "scaling in pockets." This indicates that while specific departments may benefit from automation or predictive insights, the technology is not yet integrated into the core fabric of these financial organisations.

The research identified three primary bottlenecks preventing wider adoption. Legacy infrastructure was cited as the leading constraint by 43% of respondents, as older systems often require complex workarounds to interface with modern AI tools. Data management followed at 36%, with institutions struggling against departmental silos and misaligned governance frameworks. Finally, 21% of respondents pointed to a lack of clear leadership accountability, noting that while sponsorship exists, no single executive is consistently tasked with driving AI initiatives through the entire organisation.

Naveen Bankal, Head of Banking and Financial Services for ASEAN at Cognizant, noted that while the intent to innovate is universal among regional leaders, the tone shifts significantly when discussing the current state of deployment. This suggests that Malaysian banks are caught in a transition phase, where the appetite for AI is high but the technical debt of legacy systems creates a hard ceiling for progress.

For the Malaysian consumer and SME, this "scaling in pockets" approach explains why digital banking experiences remain inconsistent. While a bank might offer a seamless AI-powered chatbot or credit scoring tool in one division, the lack of enterprise-wide integration often results in fragmented customer service experiences and slower loan processing times. For investors, this reveals that the efficiency gains promised by AI are not yet reflected on balance sheets, as banks continue to sink resources into patchwork fixes rather than holistic digital transformation.

This technical stagnation occurs against a backdrop of strong national economic performance, with real GDP growing at 6.0% year-on-year in the latest quarter. Despite this growth, banks face the dual challenge of needing to maintain operational efficiency while navigating a competitive landscape that demands rapid innovation. With headline inflation at 1.8% and unemployment hovering at 3.0%, banks are under pressure to optimize costs through technology, yet the current failure to scale AI suggests that these cost-saving measures may be harder to achieve than originally anticipated.

The reliance on legacy systems highlights a deeper structural issue within Malaysia’s financial sector, where long-standing IT architectures are now serving as a barrier to competitive evolution. As banks look to the future, they must move beyond siloed experimentation and address the core infrastructure gaps identified in the report. Whether this requires a wholesale replacement of core systems or a more aggressive middle-ware strategy remains to be seen.

The industry must now watch whether banks will prioritize capital expenditure on core system modernization or continue to favor short-term, department-specific AI projects. It remains unconfirmed which specific institutions have managed to join the 24% of banks successfully scaling AI, or what the exact financial cost of these current implementation failures is for the sector at large.