We sat down with Sean Kenny, our Financial Services Lead, to discuss the trends shaping the sector, what is top of mind for leaders right now, and where the biggest data and AI priorities are emerging as the year progresses.
From the shift from AI experimentation into production, to growing regulatory scrutiny, customer expectations, legacy modernisation and the pressure to prove ROI, financial services firms are balancing transformation with control. In this Q&A, Sean shares what he is hearing from the market, where investment is still flowing, and what separates the organisations making real progress from those still stuck in planning mode.
Q: What are the biggest trends shaping financial services right now, and which of those are having the biggest impact on data and AI priorities?
Sean: My view on the biggest trends probably reflects what different stakeholders across these businesses are prioritising now, whether that is Chief Data Officers, Chief Information Officers or Chief Financial Officers.
Trend 1 – Experimentation to production
The biggest one is the move from experimentation to production, or enterprise capability. For the last few years, AI has often been seen as a side project, with people experimenting in pockets of the business. Now, leadership teams and the C-suite are expecting real, measurable outcomes at scale.
That changes the conversation. It becomes about industrialising AI delivery. We talk a lot about MLOps, and I have even heard the phrase LLMOps governance recently. There is now a much stronger focus on model risk management, oversight, and identifying the highest value use cases rather than novelty. A lot of AI activity has been happening almost as a hobby inside desks and teams, but the focus now is on what will drive the most value and how it integrates into core business processes. How does it help firms move faster, operate better, and reduce cost? That is where the shift from pilots into production platforms becomes so important. GenAI is accelerating interest, but firms are increasingly looking at controlled deployments over speed.
Trend 2 – Regulatory scrutiny
Another major trend is regulatory scrutiny. We do a lot in remediation and regulatory-led programmes, and there is growing focus across financial services on operational resilience, customer outcomes and treatment, financial crime, and AI risk. That is pushing responsible AI, model governance, and data transparency much higher up the agenda.
From a data and AI perspective, that means explainability, auditability, end-to-end lineage, governance, and making sure there is documentation, evidence and a trail for regulators when questions come. I do not think the old model of receiving a regulatory request and taking ten days to come back with an answer is going to stay acceptable in a world where firms increasingly expect to be able to talk to their data. Bias and fairness monitoring is also hugely important in financial services.
Trend 3 – The continual shift of customer expectations
Customer expectation is another big trend. I use Monzo as my main bank, not just as a side account, and that shapes what I expect from financial services. I expect the best digital experience and real-time, personalised interactions. It would be hard to go back to a traditional experience that does not offer that.
That is driving demand for real-time data capabilities, customer 360 views, personalisation and advanced analytics. Firms are looking at how they retain and cross-sell products more effectively, and how data becomes central to the customer experience rather than something used only for reporting.
Trend 4 – Legacy modernisation
Then there is legacy modernisation and cloud adoption. A lot of firms are still constrained by decades-old core systems and fragmented architectures. The challenge is how they modernise to unlock data value while also building the right foundation for AI. That is why lakehouse and Fabric architectures, API-driven integration, and decommissioning expensive siloed systems remain such important priorities.
Q: What are the biggest data and AI challenges facing financial services firms right now?
Sean:
Challenge 1 – Data Foundations
There is definitely some overlap with the trends, but for me the number one challenge is still foundations.
Despite years of transformation programmes and millions of pounds being spent, most of the organisations I speak to still suffer from fragmented legacy estates, poor data quality, inconsistent definitions across the business, and slow access to trusted data. That creates a real problem. You can ask one department a question, ask another department the same question, and get two different answers. And when data platform rollouts move too slowly, parts of the organisation often start building their own workarounds, which only adds to the problem.
Challenge 2 – Pilot to production
Another major challenge is moving AI from pilots into production. There is a huge amount of experimentation happening, but very few organisations have actually scaled it across the enterprise. Most of the time that is not down to a lack of technical competence. The barriers are more often governance, risk appetite, operating model, and how firms prove value from what can initially be a heavy investment.
Linked to that is responsible AI. Firms need to be able to evidence explainability, fairness, auditability and control around their models, especially in areas like credit decisions, fraud, anti-money laundering, pricing and customer outcomes. If those tools are deployed without the right governance around them, the impact on customers can be significant.
There is also a real skills gap. There are not exactly huge numbers of people with deep experience of deploying AI models in banks. Organisations need advanced analytics skills, but those skills also need to be coupled with financial services domain expertise, regulatory understanding, and examples of enterprise-grade delivery. That combination is hard to find.
Challenge 3 – Proving outcomes and ROI
And then there is the ongoing pressure to prove ROI. Boards want clear commercial outcomes. In the current geopolitical environment, if firms are making significant investments, they need confidence in the return. That means use cases need demonstrable, measurable value, whether that is revenue uplift, cost reduction, risk mitigation or regulatory compliance.
Q: Even with those challenges, where are businesses still investing?
Sean:
Investment 1 – Strengthening core data platforms
Number one is strengthening core data platforms. That is still the base for scalable analytics and AI.
After that, firms are investing in AI that can show tangible business impact, whether that is revenue uplift, lower cost, reduced risk or stronger compliance outcomes. Governance controls and better data management are also high on the agenda, because firms know they cannot scale AI safely without quality, lineage, governance and the right frameworks in place.
Investment 2 – Automation everywhere
Automation is another big one. We talk a lot at Dufrain about automation everywhere as a core concept. Banks are still heavily reliant on spreadsheet-driven workflows in many areas, so there is a real opportunity to use automation more effectively to drive efficiency gains.
Investment 3 – Customer obsessed digital transformation
Customer-centric digital transformation is still a clear area of focus too, alongside the role of partner ecosystems. A lot of firms are thinking carefully about the right mix of cloud, technology and delivery partners to support the investments they want to make over the next 12 to 18 months.
Q: What separates the organisations making real progress with data and AI from those still stuck in planning mode?
Sean:
Alignment with business priorities.
For me, the biggest thing is whether the work is linked to what the C-suite is actually trying to achieve. If it is not supporting the top business priorities, then you are wasting your time.
The organisations making progress start with clear business outcomes rather than technology-led initiatives. If you start with the technology, you do not get anywhere quickly because you cannot demonstrate value to the business
Executive sponsorship
Executive sponsorship from the outset is also critical. I was speaking to a customer recently who said that, at first, AI felt like a very closed conversation. Now, as they are starting to understand the benefits, it feels like they are leaning against a slightly ajar door. Getting that buy-in early and aligning around what is actually possible is really important.
The pragmatic approach
There also needs to be a pragmatic mindset. One of the best lines I heard recently was, “feel the fear and do it anyway.” One of the big barriers with AI adoption is that organisations can become so concerned about risk that they end up doing nothing at all. The firms that move forward are the ones that can enable innovation while still maintaining control.
Back to basics
Strong data foundations still matter here too. The organisations making real progress are the ones investing properly in data quality, ownership and integration.
Q: What excites you most about working in financial services right now?
Sean:
What excites me is that financial services feels like it is at a real inflection point.
For years, firms have been investing in the foundations, and now they are beginning to see how those foundations can support transformation at scale. At the same time, the pace of change across AI and the wider technology landscape means firms have a much bigger opportunity to move beyond experimentation and start delivering real impact.
Top 3 takeaways from this conversation
- Financial services has moved beyond AI experimentation
The conversation is shifting from isolated pilots to enterprise-scale delivery, with firms under pressure to show measurable outcomes, stronger governance and clearer value. - Strong data foundations still make the difference
Legacy estates, fragmented architectures, poor data quality and inconsistent definitions continue to slow progress. The firms that invest in quality, governance, lineage and integration are in a much stronger position to scale AI safely. - The firms making progress are connecting data and AI to real business priorities
The winners are not chasing technology for technology’s sake. They are aligning investment to outcomes the business actually cares about, from customer experience and efficiency to resilience, risk reduction and ROI.
If you’re exploring how to move from ambition to action, or want to hear more perspectives from leaders across the industry, our AI Beyond Blind Trust webinar series brings together voices from financial services, industry regulators and technology experts to discuss what responsible, well‑governed AI looks like in practice.