The financial services industry is no stranger to change, but modernisation remains a complex and ongoing challenge. Banks worldwide are grappling with balancing legacy systems with new technology, staying agile in a fast-changing environment, and meeting customer expectations in an increasingly digital world.

I had the pleasure of moderating a “Power Panel” at Finovate Europe in London last week, where Thea Loch from Lloyds Banking Group, Matthijs van Voorst from ABN AMRO, Yuri Gavrilin from UZUM, and Darlene Newman from Ivy CapTech Advisors each shared their perspectives on the topic of Banking Modernisation, during which a clearer picture emerged of what banking modernisation truly entails, the obstacles banks face, and how AI might be a game-changer in overcoming some of the challenges.

The View from Large, Established Banks

Legacy Banks Are Not as Slow as You Think

Kicking off the session with a provocative question, I asked Thea Loch if she was tired of being asked when the “dinosaurs” (legacy banks) were going to dump their mainframes and finally modernise their tech stack.

Thea challenged the assumption that large banks have been slow to modernise. While challenger banks once seemed poised to disrupt traditional banking with their speed and flexibility, she argues that the product and service gap between legacy banks and digital challengers has significantly closed over recent years.

“A decade ago, it looked like the challenger banks had a significant speed advantage, launching sleek mobile-first banking experiences while legacy banks were still dealing with complex, decades-old systems. But as challengers scaled, they encountered the same complexity that legacy banks have always had to manage,” she explained.

This questions whether the challengers have had a tech advantage in the first place. We (as consumers) have been quick to praise each new product or service offering launched by a challenger bank (oh look – XYZ Challenger has launched current accounts, let’s all celebrate!), when in fact all they have really been doing is reaching parity with what the legacy banks have been offering for a long time – albeit with a sexy user interface, which the legacy banks were quick to counter.

In other words, modernisation is not just about technology—it’s also about navigating complexity at scale. The perception that legacy systems are an anchor holding banks back is often misplaced. Many of these core banking platforms, despite being decades old, remain stable, secure, and reliable. In fact, rather than abandoning them altogether, banks are finding ways to modernise around them.

Modernisation Without Full-Scale Replacement

Matthijs van Voorst, echoed similar sentiments. He emphasised that modernisation doesn’t always mean ripping out legacy systems and starting from scratch.

“There’s often an assumption that innovation requires deep integrations or heavy migrations, but that’s not always the case,” he said. “Banks can modernise by layering new capabilities on top of existing systems through APIs and partnerships.”

ABN AMRO, like other major banks, has embraced agile methodologies, cloud solutions, and partnerships with fintechs to drive innovation without completely overhauling its core banking infrastructure.

However, he acknowledged that the structure of large organisations can make change slower than in a startup environment. “A bank is not a scale-up, and we have a different kind of workforce. Innovation happens, but within the framework of a large corporate entity, which has its own constraints and risk considerations.”

Matthijs pointed out that legacy infrastructure had never really been a major obstacle inhibiting innovation. In his view, the key to modernisation is to start with small pilots and proofs of concepts to learn where you might encounter issues and then solve for those problems instead of trying to modernise everything all at once.

During a recent round table dinner we held with industry leaders in Amsterdam, a comment was made that “Legacy Thinking” (mindset) was far more of an innovation inhibitor than legacy infrastructure – a sentiment that resonated with all of the panel members.

The Need for Banking Modernisation

While legacy banks are making significant strides in modernisation, the need for continued transformation is clear. This is especially evident when looking at newer, more digitally native organisations such as UZUM.

Yuri Gavrilin offered his perspective by sharing that UZUM operates what he describes as a “FinComm” platform (do I detect a new buzzword?) - a fusion of eCommerce and fintech that provides both an eComm marketplace, and associated financial services, to consumers in Uzbekistan.

Their eComm marketplace launched just 3 years ago and now claims 1 in 3 consumers in Uzbekistan as a customer. They realised that in order to facilitate financial transactions for a population that had historically transacted in person in physical markets, they needed to provide basic banking products. Unlike traditional banks, UZUM Bank was built from the ground up with digital-first banking in mind.

Yuri explained how UZUM initially partnered with one of Uzbekistan’s largest commercial banks, to launch a digital banking experience. However, integrating with an older institution presented significant challenges. The core banking system was outdated, the processing infrastructure was slow, and even the data-sharing protocols with the central bank were stuck in the past—relying on email-based data transfer protocols (yes, you read that right ….. email to transfer data!).

UZUM took the initiative. They built their own digital banking layer on top of the existing infrastructure, modernised processing capabilities, and created seamless digital onboarding for customers.

The result? A rapidly growing neobank operating within an ecosystem that merges finance and commerce. Their success demonstrates that modernisation doesn’t necessarily mean starting from scratch—it’s about finding ways to work around existing limitations and introducing new capabilities in a scalable way.

How AI Can Accelerate Modernisation?

Decoding Legacy Systems with AI

One of the biggest hurdles in modernisation is dealing with legacy code, particularly COBOL, which still underpins a vast majority of banking transactions. Darlene Newman sees AI as a solution to this problem.

“Ninety percent of banking transactions still run on COBOL. I’m obsessed with using AI to translate COBOL into plain English so that developers can understand what the code actually does,” she said.

This is a game-changer because one of the biggest barriers to modernisation is simply understanding the logic embedded in decades-old systems. Darlene pointed out that while many companies have attempted to build AI models for translating COBOL, most are still limited in their capabilities. However, recent advancements are starting to show promise (and indeed there was a startup called tweezer.io that participated in the Finovate demo sessions that does look promising).

“The real opportunity isn’t in converting COBOL to Java—it’s in extracting the business logic and refactoring it into a more modern framework,” she explained, a view that was both echoed by Yuri and latched upon by Thea.

AI and Strategic Decision-Making

Beyond legacy code, AI is playing a growing role in helping banks prioritise modernisation efforts. Darlene is actively working on AI-powered decision-making tools that help banks determine which workflows and systems to modernise first.

“It’s not just about throwing AI at everything,” she warned. “It’s about identifying which areas will have the highest impact and using AI to strategically prioritise those changes.”

She also highlighted the role of AI in enhancing customer experience, improving fraud detection, and automating routine tasks – generally freeing up human employees to focus on higher-value activities.

The Future of Banking Modernisation

The insights from the panel revealed that banking modernisation is a multi-layered challenge. It is not simply about replacing old technology with new technology. Instead, it’s about:

• Balancing Stability and Innovation: Legacy banks don’t need to discard their mainframes, but they do need to integrate modern layers that allow them to stay competitive.
• Strategic Partnerships: Collaborations between traditional banks and fintechs can accelerate modernisation without the need for full-scale replacements.
• AI as an Enabler: AI is not a magic bullet, but when used strategically, it can play a critical role in unlocking efficiencies, improving decision-making, and translating legacy systems into modern applications.
• Leadership and Organisational Structure: Perhaps the most important factor is leadership commitment. As Thea pointed out, modernisation requires strong intent and determination to push through complex changes.

The future of banking is not about choosing between legacy and modernity—it’s about finding the right balance. Banks that succeed in modernisation will be those that blend reliability with innovation, keeping the best of their existing systems while embracing new capabilities that drive better experiences for customers.

Join us on May 22 for our flagship conference, The Banking Scene Conference Brussels, where we will bring together industry leaders, innovators, and visionaries to discuss how the technological advancements, regulatory shifts, and evolving customer expectations are driving the Metamorphosis in Banking.