AI adoption is accelerating across financial services, but the pursuit of hyper-personalisation may introduce long-term risks that organisations cannot afford to ignore, says Hans Tesselaar, Executive Director, BIAN.
AI adoption is no quick fix. This was highlighted by the revelation from MIT’s State of AI in Business 2025 report that the vast majority (95%) of corporate GenAI pilot projects do not produce measurable returns in the first half-annum. However, in a fast-moving world, an organisation’s approach to the technology is now a fundamental business consideration.
With 92% of CIOs enthusiastic about harnessing AI over the next couple of years, stakeholder buy-in is crucial. Though 69% endeavour to implement ultra-customisation, questions remain about the long-term viability of this approach.
AI as a competitive advantage
With many companies still in the experimentation stage, according to a report from McKinsey, now is the time to ask questions of governance and long-term business relevance. Customers are calling for change too, according to Capgemini’s 2025 World Banking Report, most are either unsatisfied or indifferent about their card experience.
In a sector of competition, customers will choose providers who provide them with a seamless experience over those they feel ‘loyal’ to. In 2025, the UK Current Account Switch Service supported over one million switches for the third year in a row. A consistently cited factor in these account moves was the search for improved online or mobile banking.
To compete with the efficiency of neobanks, enthusiasm is building around the impact of agentic AI. In 2025, our CIO survey, conducted with IBM, revealed that 42% are already in the pilot stage with the technology and 17% were planning launches this year.
Is hyper-personalisation the key to success?
As they plan their innovation approach, the majority of CIOs (69%) are turning to ultra-customisation, keen to use tailored partner solutions or an in-house build. Unfortunately, in an ever-changing geo-political landscape, financial institutions must be prepared to constantly evolve and keep up with market expectations. With ultra-customisation, flexibility becomes less attainable.
In contrast, a technology-stack foundation that enables interoperability – a ‘plug and play approach’ – allows for a strategy that reflects both privacy regulation change, including GDPR, and the expectations of the modern consumer who frequently turns to AI to make financial decisions. For financial institutions, it is crucial to maintain resilience across all systems, which becomes more difficult as technology ages, and expensive modifications are required to keep up.
Instead, CIOs should be looking to implement transformation that can scale – but how?
Coreless banking
A successful strategy is not held back by the siloes of legacy systems, favouring the ability to be agile and support interoperability. To accommodate for future modernisation that scales, future-thinking financial institutions are turning to coreless banking, and a ‘plug and play’ approach.
To do this, our not-for-profits standards body, the Banking Industry Architecture Network (BIAN), curated its Coreless Banking concept, supported by extremely advanced AI tools, to give institutions a strong foundation for the future of banking. Providing banks with those tools to add or swap components, a flexible core will help to drive shortened adoption time for new solutions and implementation that scales, driven by microservices and open standards that keep the strategy aligned to compliance expectations.
People buy-in
You can invest in cross-department innovation but, without buy-in from employees, the project will have restricted impact and be more likely to fail. Instead, CIOs should make a concerted effort to train workers, helping them to see solutions as a crucial part of their daily workflow and spot risks before they grow. In our report with IBM, Caio Banti, Nubank’s Chief Risk Officer Brazil, highlighted that employee inclusion was fundamental to finding value in technology; “this AI-driven evolution about the way we work and use data spans across the entire enterprise, introducing specific AI risk management requirements calling for a cultural shift. A change in which every employee, from leadership to frontline staff, acts as an AI risk manager.”
Next steps
At the conclusion of our 3rd Global Banking Summit, our experts agreed that, for financial institutions, ‘the risk of doing nothing is much greater than the risk of doing something.’ Moving forward, they should prioritise a route to innovation that suits company objectives, drives compliance, and wins customers, choosing decade-long practical success instead of aligning only with the priorities of today.
To Consider
A standardized framework, like BIAN, helps to ensure successful and compliant traceable AI flows. As the core is renewed, creating an updated tech stack, it is important to identify hidden functions and processes within the code. AI can be pivotal in assisting with this discovery phase, helping to define and create the technical environment needed for a secure transition.
