Banks are moving into a more difficult stage of technology transformation as they modernise core infrastructure and data platforms to support artificial intelligence (AI), while managing the operational and security risks created by faster digitalisation.

At Temenos Regional Forum Asia Pacific 2026 in Hanoi on 25 August, discussions ranged from Vietnam's response to digital-payment fraud to progressive core modernisation, software-as-a-service (SaaS), AI-assisted technology development and how banks can translate these investments into growth.

Vietnam is an important market for Temenos. William Dale, Managing Director for Asia Pacific, said 31 banks in the country use its systems, making Vietnam its largest individual market. About 500 people attended the forum, including 300 from Vietnam.

Vietnam strengthens controls as digital fraud evolves

Nguyen Thi Thu, Deputy Director General of the Payment Department at the State Bank of Vietnam (SBV), focused on cybersecurity and fraud prevention as digital banking and payments continue to expand.

Vietnam has progressively introduced biometric authentication for online banking transactions. The SBV is also developing SIMO, a central system through which banks and payment intermediaries share information on accounts and payment identifiers suspected of involvement in fraud.

The model allows information identified by one institution to be used by others to warn customers or subject suspicious transfers to additional checks. It moves part of fraud detection from individual banks towards shared intelligence across the payments system.

The system is already operating at scale. According to SBV data reported in July, SIMO had issued warnings to about 4.6 million customers by the end of June, prompting more than 1.5 million to pause or cancel transactions worth more than VND 5.2 trillion ($198 million). The banking sector had also completed biometric checks on more than 162.9 million individual customer records and 2.6 million organisational records, covering all accounts generating transactions through digital channels.

Vietnam is also introducing an additional safeguard for higher-risk online transfers. Under State Bank of Vietnam requirements to be implemented before 1 March 2027, individual customers will be able to set a transaction threshold and a waiting period of at least 24 hours for domestic online transfers. The waiting period applies to transfers at or above the selected threshold to beneficiary accounts that have not received an amount of that size from the customer during the preceding year. Where customers do not set their own parameters, banks will apply a default threshold of VND 400 million ($15,200) and a minimum 24-hour waiting period.

These measures reflect the changing risks created by real-time digital banking. Faster payments improve convenience, but they also reduce the time available to identify and stop fraud. Biometrics, shared fraud intelligence and, in some cases, delayed settlement provide different avenues of intervening before funds become irrecoverable.

Banks face a different modernisation question

The discussion then moved from payments infrastructure to the technology inside banks. Dale said improvements in technology were reducing implementation time and cost and argued that “the cost of inaction is now much higher than the risk of change”.

Banks nevertheless have to make a more complicated calculation. Legacy infrastructure can constrain product development, data use and resilience, but replacing or modernising it carries substantial cost and execution risk.

Takis Spiliopoulos, Chief Executive Officer of Temenos, said banks were increasingly breaking large core transformations into smaller programmes rather than replacing everything at once. Progressive modernisation allows individual businesses or technology domains to move at different speeds and reduces the risk concentrated in a single migration.

The business case remains important. Foo Boon Ping, President and Managing Editor of TAB Global, publisher of The Asian Banker, said during a regional banking leaders' panel that management ultimately has to deliver profitable growth and sustainable returns. Modernisation has to contribute to those outcomes rather than become an objective in itself.

With greater uncertainty around margins and interest rates, banks are looking beyond net interest income towards fee and non-interest income. Technology can help them understand customer behaviour and needs more accurately and turn transaction and engagement data into more relevant products, advice and deeper relationships.

AI puts greater pressure on core and data

Foo described the development of bank technology in two broad generations. The first digitised and connected the bank through internet and mobile banking, digital customer journeys, process automation and application programming interfaces (APIs). Banks could substantially improve the customer experience without necessarily replacing the systems underneath.

The next stage involves more fundamental changes to core systems, architecture, cloud and data. AI increases the importance of those investments because its effectiveness depends heavily on the quality and accessibility of the underlying data.

Spiliopoulos said banks he speaks to are concentrating on digital customer experience, legacy modernisation and data foundations for AI. He said banking applications also have to meet requirements around accuracy, explainability, auditability, resilience and compliance.  “The starting point is always a data problem,” he said.

This does not mean banks need to replace their cores before deploying AI. Many are already using it selectively. The constraint becomes more apparent when institutions try to extend those applications across fragmented data and complex legacy systems.

Execution remains difficult

Bank of New Zealand (BNZ) and Australia's Alex Bank illustrated different approaches. David Morgan, General Manager Technology – Core Ledger at BNZ, said major core programmes require internal expertise, experienced partners, extensive rehearsal and the willingness to make genuine go/no-go decisions before migration.

Simon Beitz, Chief Executive Officer and co-founder of Alex Bank, said the digital bank chose SaaS because it did not want to build and operate technology that did not differentiate its business.

Ganesan Sriraman, Global Head of Product Engineering at Temenos, said successful modernisation also depends on controlling scope from the outset, continuously reviewing the implementation plan and breaking transformation into smaller increments. His approach was to “be small to win big”, allowing banks to make progressive changes rather than treating modernisation as a single large programme.

William Moroney, Chief Revenue Officer at Temenos, put greater emphasis on the relationship between institutions and their technology providers. He said banks undertaking complex transformation should treat the companies involved as partners rather than conventional vendors, particularly when programmes encounter problems. He also linked the emerging use of AI to trust, arguing that its growth potential depends on how safely it can be incorporated into banking.

The approach reduces some of the infrastructure a smaller bank has to manage itself, but increases its dependence on the technology provider for resilience, upgrades and platform development. Beitz described it as a 10- to 15-year relationship rather than a conventional technology project.

Both also saw potential for AI to change how transformation programmes are executed. Beitz identified documentation, testing and customer-journey design, while Morgan pointed to requirements engineering, design, testing and resource management.

These applications can improve productivity, but it is too early to conclude that they will materially reduce the overall cost or risk of large core programmes. Integration, data migration, organisational change and cutover remain substantial parts of the work.

Trust becomes harder with AI

Noel Santiago, Banking Transformation and Digital Banking Advisor and former Head of Digital Banking at BPI, said the move towards AI agents creates a different governance problem. Agents can move outside intended boundaries as they interact and learn, making monitoring, policy controls and clear operating limits important before they are deployed in production. He also cautioned that the economics remain uncertain as usage scales and AI consumption increases.

AI also changes the customer side of the equation. Foo said customers have traditionally trusted banks to safeguard their money and information. As AI moves into customer interactions, recommendations and decisions, they also need confidence that the identity is genuine, their data is being used appropriately and the bank can explain and stand behind an AI-influenced outcome.

Banks therefore have to increase the frequency of technology change without increasing failures or weakening controls.

The experience with digital payments in Vietnam makes the point more concretely. Speed and convenience have created customer value, but also required biometrics, shared fraud intelligence and additional controls for transactions carrying greater risk.

AI presents a similar challenge. It may improve productivity, customer understanding and eventually the economics of technology transformation. But those benefits will have to be demonstrated in operating results rather than assumed from the technology itself.

For banks, the measure remains whether modernisation improves growth, customer outcomes and the ability to change while maintaining the resilience and trust expected of a financial institution.