Financial institutions across Asia face increasing pressure to modernise rapidly amid rising digital transactions, evolving customer demands, tighter regulation and ongoing innovation. Technology transformation is now a strategic differentiator. At the Asian Banker’s Heads of Technology Roundtable, held in partnership with Temenos and Red Hat, senior executives from financial institutions discussed how cloud, artificial intelligence (AI) and core technology integration drive this transformation. The session examined challenges, and the organisational and technical changes needed to build future-ready banks. Data readiness is essential for transformation Banks need to be intelligent, resilient and hyperpersonalised at scale using data for real-time decisions and risk management. Legacy systems, fragmented data and rising demands from customers and regulators pose key challenges. One executive stressed that organisational alignment and addressing structural weaknesses, rather than technology alone, remain the greatest barriers. Resistance often stems from IT teams’ long-term ownership of legacy systems. Executives agreed on the ongoing challenge of achieving a single version of truth and unified customer data across subsidiaries. One bank shared that they have established a dedicated data and AI governance function reporting directly to the board. Another emphasised the complexity, sharing that they have spent four years progressing towards unified data management, expecting another two years before completion. Developing transformation strategy The discussion emphasised that transformation should be guided by clear business needs and a defined adoption sequence aligned with long-term goals, not by blind technological leaps. A key challenge is evolving legacy cores while innovating at pace. Participants shared emerging practical models from across the region. Future readiness must align with current needs. One participant warned against chasing technology perfection in isolation, urging firms to first define business outcomes and then identify enabling technologies. One bank overhauled governance by replacing multiple overlapping committees with a single digital board chaired by the chairperson and CEO. This streamlined decision-making allowed new initiatives to move from approval to launch within weeks. Speed and time to market, rather than perfection, became the priority. Incremental modernisation of core technology to prepare for the future Executives highlighted microservices-based cores as essential for banks facing sharp transaction spikes, while stable-growth banks may rely on legacy cores augmented with layered services. One bank outlined a dual approach—cautious incremental upgrades alongside cloud-native experimentation. The executive emphasised the need to manage complexity. Its core banking system remains largely legacy and on-premise to maintain stability, while it develops a next-generation mobile banking app fully built on cloud-native architecture using Kubernetes and DevOps. This parallel approach aims to strike a balance over time. The incremental track learns from the radical track’s speed and flexibility, while the radical side appreciates the reliability and cost efficiency of traditional systems. The future, the executive suggested, lies in combining both to deliver the best solution for each use case. This enabled a shift from rigid over-provisioning to a hybrid cloud model improved efficiency—provisioning mobile platforms for two to three times capacity instead of five to 10 times, scaling with the cloud during peaks. Another executive advocated decoupling core systems from digital delivery layers. By abstracting the core through application programming interfaces (APIs), innovation can continue at market speed even as backend modernisation proceeds. Nonetheless, he cautioned that the core must eventually scale to support growth. Imrish Singh, business development director at Red Hat, observed: “The three typical challenges we see with clients are siloed data, lack of expertise and integrating two-speed IT—how you bring together modern cloud-native systems with existing infrastructure.” Where core transformation once meant high-risk replacements, banks now favour measured approaches. Zannettos (Zan) Zannettou, BSG manager-technology, Temenos explained the rising preference for modular modernisation to reduce risk and accelerate value: “Banks want to avoid the ‘big bang’ approach and bring value to customers faster through modular, low-risk implementation.” He cited a bank migrating its payment platform from on-premise to self-hosted public cloud, working closely with regulators and leveraging internal expertise. Zannettou stressed the importance of integration, supporting modern APIs and streaming while accommodating legacy systems. He added, “We offer a SaaS (Software as a Service) Accelerate service to facilitate cloud testing. Banks also demand simpler customisation. We’re embedding AI in our development platform to accelerate customisation and reduce time to market.” Banks transition to cloud while managing barriers to scale Cloud adoption is central to digital transformation but raises complex issues around architecture, control, compliance and resilience—especially for core and regulated workloads. Institutions shared that as they migrate to public cloud, legacy systems persist due to regulatory complexities, making full core modernisation a gradual process. Some adopt a hybrid cloud strategy, retaining part of their core on-premise while moving other components to the cloud to balance resilience and scalability. A digital-only bank revealed it recently completed a full-stack cloud-native core banking switch within 10 months, cutting IT costs by 40%. The executive advised banks considering core banking change to “take the leap, as the longer you wait, the harder it becomes.” An Indonesian digital bank operates two core systems—for wholesale and retail—and has developed internal composable services to bridge the two platforms, accelerating development and reducing vendor friction. Cloud infrastructure enabled scale and speed, while core systems remained on-premise for regulatory compliance. Regulatory engagement emerged as a key theme. One participant from Hong Kong detailed close collaboration with regulators during a 10-month core migration, involving real-time progress updates via a shared channel and post-launch app testing by the regulator. In the Philippines, joint efforts between banks, fintechs and regulators have fostered a progressively supportive policy environment. Imrish Singh business development director, ISV Ecosystem, FSI Asia Pacific at Red Hat highlighted two cloud priorities: “Supporting both modern and legacy virtualised IT workloads through a unified platform is now essential for scalability and control. Strategically, multi-cloud portability, and having an integrated platform that manages everything from physical infrastructure to platform management are key.” Panellists agreed that platform architecture, modularity and virtualisation are vital to future-proof technology. Avoiding vendor lock-in is critical. One participant said their systems operate as “two-way doors,” with a central platform allowing testing, reversal or adjustment without long-term constraint. Strategies to scale AI responsibly As the discussion moved to AI, participants shared how their organisations balance speed, responsibility and scale in enterprise-wide adoption. A Thai Bank described a two-pronged approach to democratising AI. From the top down, business units define strategic goals and identify AI opportunities. From the bottom up, the bank trains 20,000 employees on AI use cases and tools, promoting self-service innovation. This dual-track is supported by strong governance, aligning innovation with policies and regulations, and enforcing strict data privacy and third-party tool limits. Another participant described a “Rambo-style” AI implementation at a digital bank, prioritising rapid deployment over formalised procedures. Within three months, the bank introduced generative AI (GenAI) tools across front, middle and back-office functions, leveraging open-source models. Use cases include chatbots for customer engagement, machine learning for credit analysis, and GenAI for regulatory document searching. Rejecting silos, the bank focused on business agenda, integrated technology with product teams to deliver immediate business value. One participant noted that traditional banking systems are still designed around human intervention despite automation. Banks will likely need to redesign platforms to accommodate fully autonomous AI-driven workflows with real-time automated decision-making. One bank operates nearly 30 AI use cases on-premise, including productivity chatbots and customer service. Some processes like risk scoring and settlements are fully automated but require human oversight. The bank emphasised engineering discipline, mature continuous integration/continuous delivery (CI/CD) pipelines, evaluation frameworks and continuous training to ensure AI aligns with expert judgement. Optical character recognition (OCR) use cases have surpassed human performance. Trust and regulatory compliance in AI Rapid AI adoption must not compromise compliance. Participants stressed transformation must conform, perform and then transform. One executive noted that building trust with security teams and regulators requires providing monitored, filtered and controlled AI experimentation environments. They focus on internal safeguards and maintain tight control over AI deployment in production settings. Some institutions are building flexible AI platforms that allow multiple models—both commercial and open-source—to operate within a unified framework. These platforms offer internal teams a secure sandbox for experimentation while giving security and governance teams a single point of control. This approach prevents shadow AI use and ensures regulatory compliance. A shared concern is the risk of AI hallucinations. Participants warned that while AI can boost productivity, it must remain a co-pilot, not a pilot. Human oversight is non-negotiable. Training staff to understand AI limitations and verify outputs is essential for responsible deployment. Executives agreed AI should support not replace, human decisions. Logical frameworks, policy alignment and continuous education are vital for effective AI use. Internal alignment is key. Centres of Excellence bridge IT, compliance, security and business, educating teams and ensuring regulatory adherence. AI innovation is accelerating. One bank recently deployed a fully automated back-office process unimaginable months prior. Panellists warned AI adoption is no longer optional: “Bad actors also use AI.” Collaboration with regulators and law enforcement is vital to counter AI-driven cyber threats. Cost concerns were also raised. Singh emphasised trust, transparency and legal indemnification as critical pillars for AI adoption in regulated banks. He shared that their platform supports custom training on enterprise data to meet compliance while maintaining flexibility and portability. Zannettou shared that they embed AI in core banking functions, including explainable AI for credit approvals and GenAI to accelerate product configuration. GenAI also enables efficient querying of disparate data for marketing and faster product to market and a low-code, no-code approach. To integrate third-party AI, Temenos offers secure REST APIs and real-time streaming. It’s marketplace hosts validated partners with adapters for rapid deployment. Data security is critical. “On-premise models ensure security. Cloud models can also be secure, but banks worry about data sharing due to regulatory constraints,” said Zannettou. Key takeaways The discussion underscored the importance of a structured approach to strategic transformation. Executives highlighted the need for cautious and secure adoption of public cloud, balancing scalability with security and regulatory compliance. Innovation must proceed alongside maintaining legacy system stability, with modernisation strategies ranging from full core replacements to modular, low-disruption approaches. AI adoption requires strong top-down leadership combined with bottom-up engagement, supported by robust governance. The industry must move from treating data as a cost to recognising it as a core asset that drives customer value and operational efficiency. Data maturity varies across markets, with many still lacking the longitudinal data necessary for advanced AI, making partnerships essential to keep pace with technological advances. Multi-cloud and AI-first strategies are becoming standard, but concentration risks mean banks must select partners carefully and maintain clear strategic focus. Ultimately, the greatest challenge remains “using AI to keep up with AI.” This fast-evolving landscape calls for close collaboration with regulators and the development of industry-wide solutions to safeguard security and build trust.
