Ten years ago, I spoke to Global Finance about innovation in banking and payments. At the time, APIs and blockchain were being positioned as technologies capable of transforming financial services. Banks were investing in accelerators, fintech was challenging established business models and corporate customers were demanding greater flexibility and transparency.
A decade later, the technology conversation has moved on considerably. Cloud infrastructure has matured. APIs have become part of the fabric of financial services. Real-time payments have expanded. And now AI promises another significant shift in how financial institutions and businesses operate.
Yet one question feels remarkably familiar: are we genuinely transforming how things work, or are we using new technology to improve processes that should have been rethought altogether?
Innovation is not the same as digitisation
Back then, one of the criticisms levelled at financial services was that much of what we called innovation amounted to taking an existing process and making it digital. I described existing solutions being repackaged and themed as “digital”.
We’ve made enormous progress since, but the distinction still matters.
Replacing a spreadsheet with a dashboard is useful. Automating a manual process can deliver significant efficiencies. Applying AI to analyse information that previously required hours of human effort can create enormous value.
But none of those things automatically constitutes transformation.
The more important question is whether technology allows us to rethink the process itself: why it exists, where friction has accumulated, what customers actually need and whether there is now a fundamentally better way to achieve the outcome.
That distinction becomes particularly important with AI. If we simply apply AI to inefficient workflows, we risk creating more sophisticated versions of the processes we already have.
The barriers aren’t always technological
Another observation I made at the time was that banks wanted to become more disruptive and agile but were understandably constrained by compliance and regulatory considerations. Corporate customers wanted flexibility, transparency and innovation, while institutions had to balance those expectations against the responsibilities that come with operating critical financial infrastructure.
That tension hasn’t disappeared either.
Technology can move extraordinarily quickly. Organisations cannot always move at the same pace — nor should they.
Payments sit inside highly regulated, interconnected environments where changes can affect customers, counterparties, financial crime controls, data, liquidity and operational resilience. The challenge isn’t simply whether a new technology works. It is whether organisations can introduce it with the governance, oversight and accountability required to trust it.
That is particularly relevant as businesses move from experimenting with generative AI towards embedding AI within operational workflows.
Start with the problem, not the technology
Perhaps the biggest lesson from the past decade is that predicting which technology will change an industry is less useful than understanding which problems actually need solving.
In 2016, blockchain was expected to reshape large parts of transaction banking. I argued at the time that its potential depended on government and central-bank endorsement and, critically, standards. APIs ultimately became far more pervasive because they solved immediate problems around connectivity and access to information.
AI will follow its own path.
Some use cases will prove transformative. Others will disappear. And many of the greatest gains may come from applications that currently attract far less attention than the technology itself.
For payments, that means starting with the operational and commercial problem. Where is complexity creating unnecessary cost? Where is poor visibility affecting performance? Where are teams spending time on repetitive analysis? Where is valuable knowledge fragmented across systems and people? And where could better intelligence materially improve an outcome?
Only then should we ask what role AI can play.
Ten years on, perhaps the breakthrough isn’t a technology
Financial services has no shortage of technology.
What remains harder is bringing together technology, deep domain expertise, organisational readiness and appropriate governance around a clearly defined business problem.
That was true when we were talking about APIs and blockchain.
It’s even more important now.
Because the organisations that gain most from the next wave of innovation probably won’t be those that adopt AI first. They’ll be the ones that are clearest about where it genuinely makes the business better.
Originally inspired by Andrew Doukanaris’ contribution to Global Finance’s 2016 feature, Building Toward A Breakthrough.
Read the original Global Finance article