Temenos Community Forum (TCF) 2026 in Copenhagen opened with a more practical question for banks than whether they should modernise. The issue now is how they modernise without destabilising the systems that already support customers, branches, payments, wealth, lending and regulatory reporting every day.

Takis Spiliopoulos, chief executive officer at Temenos, described an industry operating under growing pressure to maintain “absolute stability” while continuing to evolve. He linked that pressure directly to artificial intelligence (AI), arguing that banks are no longer asking if they should modernise, but how they do so while limiting disruption to customers and operations.

The plenary repeatedly returned to the tension between maintaining stable production systems and introducing new capabilities quickly enough to support AI, real-time processing and increasingly digital customer expectations.

Spiliopoulos argued that banks are moving away from large “big bang” transformation programmes towards more progressive approaches, introducing capabilities incrementally while attempting to reduce disruption. But the discussions also made clear that progressive modernisation does not necessarily reduce complexity. In many cases, it redistributes it across longer timelines, broader integration layers and more continuous upgrade cycles.

Will Moroney, chief revenue officer at Temenos, described a noticeable shift in customer behaviour. Smaller banks and digital institutions had already modernised large parts of their technology stacks over the past decade, but he said the momentum is now “going up the tier of banks”, with Tier 1 and Tier 2 institutions increasingly bringing core modernisation back onto the strategic agenda.

Moroney argued that large incumbent banks are constrained not only by ageing systems, but also by legacy processes, organisational structures and long-established ways of working. As AI becomes a larger operational priority, those constraints are becoming more visible.

At the same time, the session reflected uncertainty around how easily very large banks can sustain continuous transformation models over long periods. Several discussions suggested that maintaining shorter upgrade cycles, integrating AI safely and modernising progressively may ultimately require a level of governance, delivery consistency and organisational discipline that many banks are still developing.

Tier 1 banks reconsider the cost of delayed modernisation

One senior executive from a large international private banking group described how the institution came to view regular upgrades less as a technical requirement and more as part of maintaining delivery speed, stability and predictable cost management.

The executive said the bank realised shortly after going live that it needed to remain close to the upgrade cycle rather than allowing changes to accumulate over extended periods. The bank supported that approach with automation across regression testing, build pipelines and provisioning.

Developers could request environments and receive them within minutes, reducing dependency on specific individuals and improving development speed. But the executive also made clear that maintaining that discipline required continuous attention rather than periodic intervention.

The example highlighted an issue that surfaced repeatedly throughout the session: the cost of staying current may be lower than the disruption associated with delayed upgrades and larger transformation exercises later. At the same time, continuously maintaining modern systems also requires sustained investment, automation capability and internal coordination.

Barb Morgan, chief product and technology officer at Temenos, linked that reality directly to the company’s product strategy. She argued that banks increasingly want to modernise “capability by capability”, allowing specific domains to evolve independently without forcing wholesale replacement of surrounding systems.

That logic sat behind Temenos’ announcement that retail deposits and lending are now available as composable solutions. Morgan positioned composability as a way for banks to isolate and modernise particular domains more progressively. But composability itself also introduces new demands around orchestration, governance and lifecycle management across increasingly interconnected systems.

In a later media discussion, Moroney said some large banks are increasingly exploring sidecar or carve-out approaches, keeping legacy systems in place while launching selected products or capabilities on newer platforms. He argued that many institutions now prefer shorter implementation cycles and cleaner upgrade paths rather than waiting for large-scale replacement programmes to conclude. Morgan added that banks increasingly want less customisation within the core itself, while preserving flexibility and personalisation at the customer level.

Shyam Gopal Rajagopalan, head of operations platform at Raiffeisen Bank International, described composability as a way for banks to reduce the disruption traditionally associated with large-scale core replacement programmes. He argued that many institutions are moving away from “all or nothing” transformation approaches in favour of progressively modernising specific capabilities, although doing so still requires banks to manage ecosystem complexity, integration and operating model change carefully.

SaaS changes the operating trade-offs

David Furlong, chief technology officer at Questbank, a new digital bank launched by Questrade Financial Group, approached the issue from a different angle. Questrade, which Moroney described as Canada’s largest online brokerage firm, went live on a software-as-a-service (SaaS) platform in nine months as it launched the digital bank. Furlong said the decision was partly driven by resource allocation. The organisation did not want large parts of its technology team tied up maintaining the underlying platform and associated support work.

Instead, the bank wanted more resources focused on customer-facing development and speed-to-market initiatives. But Furlong also acknowledged that adopting SaaS required the business itself to simplify. Questbank changed pricing structures, policies and procedures rather than heavily customising the platform.

That trade-off became one of the more important themes in the session. SaaS may reduce some infrastructure and upgrade burdens, but it can also require banks to adapt operating models, governance structures and business processes to fit more standardised platforms.

Furlong described a newer banking business deliberately simplifying operations to avoid dedicating scarce technology capacity to running the core. Both approaches attempted to solve the same problem differently: how to prevent the core platform from absorbing disproportionate organisational attention.

The discussions also highlighted how SaaS and progressive modernisation models shift part of the dependency away from the bank itself towards external vendors and delivery partners. That increases the importance of governance, transparency and long-term alignment between institutions and providers.

Marnix Tummers, IT director of wealth management at ABN AMRO, described how the bank used a more harmonised platform strategy across markets including the Netherlands, Belgium, Germany and France as it integrated a recently acquired German private banking business onto a modern wealth platform within compressed timelines. He linked the bank’s technology priorities partly to changing customer expectations around digital access, personalisation and intergenerational wealth transfer.

AI exposes pressure points in data and operating models

Jiří Kacerovský, chief data officer and head of Data Centre of Excellence at Komerční banka, brought the discussion into the data layer, where many of the pressures surrounding AI are becoming more visible. He argued that core transformation programmes can either improve a bank’s data architecture or add another layer of fragmentation, depending on how the institution approaches the transition.

Kacerovský described two possible approaches. A bank can integrate a new core platform simply as another data source within an already complex architecture, or it can use the transformation programme to rethink governance, ownership and data structures more broadly. Komerční banka chose the second approach, using the programme to introduce data product concepts, clearer ownership structures and more manageable delivery components.

But Kacerovský also acknowledged that establishing trusted data structures across large organisations remains difficult, particularly where institutions operate across multiple legacy systems and business structures.

AI readiness therefore depends less on models alone than on whether banks can maintain usable, governed and reliable foundations underneath them. Morgan reflected a similar position from the product perspective. She described Temenos’ AI strategy less as a separate overlay and more as intelligence embedded into workflows, operations and banking processes already running within the platform.

AI may expose weaknesses in legacy banking systems faster than many institutions can realistically modernise them. Fragmented data structures, inconsistent workflows and ageing processes all become harder to manage once banks attempt to scale AI across production systems.

Delivery discipline becomes part of the transformation challenge

Implementation and delivery also emerged as major themes throughout the session. Moroney described large transformation programmes as ecosystems involving the bank, the core platform, surrounding systems, implementation partners and operational teams. He noted that banks may spend 18 to 24 months evaluating projects before contracts are signed, followed by years of migration, implementation and upgrade activity.

That longer lifecycle increasingly places pressure not only on technology platforms, but also on governance structures, programme management and delivery consistency over extended periods. Moroney said Temenos has expanded its hybrid delivery model, customer assurance programme and global delivery capabilities in response to the growing complexity of implementation programmes.

AI is also becoming part of that delivery process. Moroney described how Temenos is developing micro-agents within its delivery operations to extract business logic from older systems, automate testing and reduce onboarding and implementation timelines.

But AI may not necessarily simplify transformation immediately. In some cases, it may increase pressure on banks to modernise systems, improve governance and maintain current platforms faster than existing operating models comfortably allow.

Christine Huberty, chief information officer at Banque Internationale à Luxembourg, described how the bank modernised payments infrastructure while simultaneously transforming its broader banking architecture, including strengthening application programming interface connectivity and adapting to instant payments and evolving regulatory requirements. She said the bank processes approximately two million payments daily with straight-through processing rates of around 99.6%.

The strongest message emerging from TCF 2026 was therefore not that banks have solved the problem of core transformation. Rather, banks appear to be searching for less disruptive ways to manage increasingly difficult operating conditions as AI, real-time processing and composable architectures become more deeply embedded into banking operations.

The banks’ perspectives illustrated different responses to that challenge. One large private banking institution focused on upgrading discipline and automation. Questbank simplified operations to support a SaaS model and preserve technology capacity. Komerční banka used core transformation as part of a broader attempt to modernise data governance and ownership.

The discussions at TCF 2026 suggested that the longer-term challenge for large banks may no longer be replacing the core itself, but sustaining the governance, delivery capability and organisational discipline required to keep continuously evolving systems under control as AI becomes more deeply embedded across banking operations.