The buzz at Money20/20 Europe this year was palpably different. While last year’s conversations were dominated by the hypothetical potential of new technologies, 2025 has seen the dialogue shift decisively towards practical application, measurable ROI, and the foundational work required to make innovation a reality.
Nowhere was this clearer than in the discussions around Artificial Intelligence and digital assets. Bobsguide sat down with Michael Levens, who leads the Data, Tech, AI, and Payments practices at consulting and technology firm Delta Capita, to discuss the key takeaways from the event and the evolving landscape for financial institutions.
Delta Capita, which provides consulting, managed services, and proprietary fintech products, operates at the heart of this transformation. The firm often acquires technology from top-tier banks like Citigroup, LSEG, and HSBC, revamping it to create mutualized, market-wide solutions aimed at reinventing the financial value chain.
“Last year was very much hypothetical AI stuff,” Levens began. “Now we’re talking about agentic AI and what the real use cases are in the financial services industry.”
The AI Evolution Requires a Solid Data Foundation
The evolution from generative AI—focused on the value of the content it creates—to agentic AI, which can perform autonomous tasks, represents a significant leap. However, Levens cautioned that this progress is entirely dependent on a critical, often-overlooked prerequisite: data quality.
“Clients came to us and said, ‘We want to do AI.’ We dug a little bit deeper, and we’d ask, ‘How good is your data?’” Levens shared. “Often, it came back to, well, our data isn’t good enough.”
This challenge is the unseen handbrake on innovation for many established institutions. According to Levens, the financial services organizations poised to accelerate and truly leverage AI are those that have invested the last decade in their data governance, quality, and lineage. Without a clean, well-structured data infrastructure, attempts to implement advanced AI are fraught with risk.
“If the data isn’t good enough, they’re going to be just building on top of trouble,” he noted. This reality forces a strategic dilemma: institutions can’t afford to wait until their ‘spaghetti architecture’ is perfectly sorted, yet they must be prepared. This explains why a significant portion of Delta Capita’s consulting work involves helping banks improve their underlying data infrastructure before they can scale their AI ambitions.
A core use case Delta Capita has developed in response to this need is an AI risk assessment toolkit, designed to help organizations rapidly determine if their AI policies and strategies are fit for purpose and compliant with emerging regulations like the EU AI Act.
Stablecoins: Solving the Cross-Border Conundrum
Beyond AI, the other topic dominating high-level discussions was stablecoins. Once on the periphery of institutional finance, stablecoins are now being seriously evaluated as a practical solution to one of the industry’s most persistent challenges.
“The whole industry has been looking at solving the cross-border conundrum of frictionless, real-time payments,” Levens explained. “Can stablecoin really help with that? I think there is definitely use cases.”
Unlike cryptocurrencies, which have rarely entered top-tier banking discussions due to a lack of regulation, stablecoins are gaining traction precisely because regulators like the Bank of England and the US Federal Reserve are building frameworks around them. Being pegged to a fiat currency provides an additional “ingredient for success.”
Levens believes initial adoption will be faster in commercial payments, where participants have a greater understanding of payment rails and the associated costs. For retail, the path is more complex.
“Are we going to be educating our retail customers around what a stablecoin is?” he posited. The ideal future state, he suggests, is one where the end-user doesn’t need to be educated. “You just want to make a payment in the easiest or fastest way… The organization you’re paying through will work out whether it’s through stablecoin or some other rails.”
Following a pattern similar to AI’s adoption curve last year, Levens predicts that H1 2025 has been exploratory for stablecoins, with concrete use cases starting to emerge in H2, especially as the industry gets to grips with new regulations.
Build, Buy, or Partner?
The challenge of implementing these new technologies feeds into a larger strategic question for financial institutions: should they build capabilities internally, buy them off the shelf, or partner with fintechs?
Levens pointed to the example of Monzo, which famously built its own core infrastructure, a feat that is difficult to replicate for an incumbent bank burdened with decades of technical debt.
“If you’re a large bank of 50 years, or even 20 years, you’ve got so much technical debt. The starting point is so different from a digital bank like Monzo or Revolut,” he said. This reality drives many large institutions to partner, though some remain wary of being “hamstrung by vendors.”
This complex dynamic underpins Delta Capita’s strategy of acquiring and mutualizing non-differentiating technology from banks, aiming to create shared services that benefit the entire industry—a model the firm plans to expand aggressively over the next 12 months.
As the financial world moves from hype to implementation, the message from Money20/20 is clear: success will belong not just to the most innovative, but to those who build that innovation on a solid foundation of data, strategy, and regulatory awareness.
