Financial institutions across Malaysia are integrating artificial intelligence into their daily operations at an accelerating pace, yet substantial doubts persist about entrusting the technology with consequential strategic decisions. A comprehensive study by the Asian Institute of Chartered Bankers (AICB), conducted in partnership with Ecosystm and the AICB Chief Risk Officers' Forum, paints a picture of an industry in transition—keen to harness AI's potential but wrestling with fundamental questions about trustworthiness and oversight.
The research, unveiled during AICB's 4th Malaysian Banking Conference and 2nd Bank Audit Conference in Kuala Lumpur, surveyed 87 senior leaders from commercial, digital, and Islamic banks alongside development financial institutions. The findings reveal a sector applying AI to increasingly sophisticated functions, from streamlining customer verification protocols to identifying suspicious transaction patterns and strengthening defences against money laundering and terrorist financing. Employee productivity enhancements represent another growing application. Yet this expansion into critical domains has triggered a parallel anxiety: just one-quarter of respondents expressed sufficient confidence in AI-generated outputs to base major business decisions upon them.
This confidence gap underscores a deeper institutional challenge. Edward Ling, AICB chief executive, reframed the conversation away from whether AI belongs in banking toward whether institutions possess the ethical foundation, governance structures, and professional acumen to deploy it responsibly. The question moves beyond technological capability to organisational maturity—whether decision-makers can account for how AI recommendations might affect customers, reshape risk profiles, and influence competitive standing. For Malaysian banks operating in an increasingly scrutinised regulatory environment, this distinction carries weight.
The complexity inherent in AI systems themselves compounds the challenge. Chong Han Hwee, chair of the AICB Chief Risk Officers' Forum and group chief risk officer at RHB Malaysia, highlighted how AI-related risks transcend the algorithms themselves. Problems emerge throughout entire operational ecosystems, spanning data quality issues, how humans interpret and act upon AI suggestions, and the downstream consequences of those decisions. This multifaceted risk landscape evolves continuously, making static safeguards inadequate. Banks cannot simply implement AI and assume stable performance; they must monitor and adapt governance frameworks as the technology and business context shift.
The study identified stark readiness disparities across Malaysia's banking sector. Nearly half—44 per cent—remain in a developmental phase, having moved past initial experimentation yet still wrestling with fragmented data capabilities, uneven skill distribution, and operating models poorly aligned for AI integration. Only 15 per cent have attained an established level of readiness, while a mere 2 per cent occupy the advanced tier where AI is seamlessly woven into decision-making processes and delivers competitive advantages. This distribution suggests most Malaysian banks face a multi-year journey before AI becomes foundational to their business model.
Strategic alignment represents a critical bottleneck. Just over one-quarter of institutions maintain clearly articulated strategies linking AI initiatives to business objectives. Conversely, 44 per cent are already constructing custom AI solutions—a figure that raises concerns about proliferating fragmented projects that resist scaling or replication across business units. Without overarching strategic frameworks, banks risk creating isolated AI pockets that consume resources without generating enterprise-wide value. This scattered approach also complicates efforts to establish consistent governance and controls.
Human capital shortages amplify these structural weaknesses. An overwhelming 79 per cent of respondent institutions report inadequate supplies of specialised AI technical talent. More troubling still, only one-fifth actively cultivate AI-literacy across their workforces, suggesting that most banking professionals lack foundational understanding of how AI systems function and where they introduce risks. This gap between tool deployment and workforce capability creates vulnerabilities: employees may misuse AI outputs, fail to recognise when algorithms produce flawed recommendations, or lack the conceptual grounding to contribute meaningfully to governance discussions.
Governance deficiencies represent perhaps the most pressing concern revealed by the research. Roughly half of Malaysian banks and DFIs still rely on informal, ad hoc governance rather than systematic, risk-proportionate frameworks to evaluate different AI applications. Only one-third have established structured governance and model risk management processes, while barely more than a quarter apply formal risk tiering—a foundational practice for tailoring oversight intensity to the riskiness of specific use cases. These governance gaps are particularly concerning given the potential impact of AI failures on customer trust, regulatory compliance, and financial stability.
Sash Mukherjee, vice-president of industry insights at Ecosystm, articulated an emerging consensus: as AI ventures into higher-stakes decision-making, financial institutions are demanding greater transparency around model risk management, algorithm explainability, third-party AI tool verification, and data governance disciplines. The appetite for regulatory clarity is evident. Yet Mukherjee cautioned that regulation alone will not suffice; policymakers cannot legislate fast enough to keep pace with rapidly evolving AI capabilities. Meaningful progress requires sustained, structured dialogue between the banking industry and regulators to ensure governance frameworks mature in tandem with technological advancement.
For Malaysian readers and policymakers, the report signals both opportunity and urgency. The banking sector's willingness to experiment with AI reflects confidence in its potential to enhance customer service, operational efficiency, and risk detection. Simultaneously, widespread governance gaps and skill deficits mean many institutions lack the infrastructure to deploy AI safely at scale. Bank Negara Malaysia and other regulatory bodies must intensify efforts to build industry consensus around governance best practices while encouraging capacity-building initiatives. Individual banks, meanwhile, should prioritise establishing clear AI strategies, investing in staff development, and implementing systematic governance before expanding AI applications into riskier domains. The next phase of Malaysia's AI adoption in banking will determine whether the sector harnesses the technology responsibly or stumbles into preventable crises.
