AI in banking is well past the proof-of-concept stage. Lloyds Banking Group alone has reported more than 50 generative AI use cases and 200 active use cases in the year gone by, delivering over £50 million in value. And with 1,500 new tech and data specialists hired in 2025 and huge targets set for 2026, this reflects a sector-wide shift.
As such, institutions are now embedding AI into multi-year investment strategies, building dedicated centres of excellence, and reskilling workforces at scale.
So, the efficiency gains are real and measurable. But a question the current wave of investment hasn’t meaningfully answered is what AI does for the parts of banking that depend on coordination between institutions, not just internal efficiency.
This article will explore what AI is being used for in banking already, the gaps it hasn’t plugged, and whether it’s capable of doing so.
Where AI in Banking Is Delivering Results
The clearest gains from AI in banking are concentrated within individual institutions.
Banks are automating back-office processes, augmenting risk management, and testing customer-facing tools in controlled environments. They’re processing backlogs that previously took weeks in days, too, and risk functions are operating with greater speed and precision than ever.
Lloyds’ approach in particular illustrates the model many larger institutions are following. Capability is built centrally, then distributed across the business. Workforce roles are changing rather than disappearing, and traditional functions such as risk management, project delivery, and workforce strategy are all developing AI-linked variants. So, the technology is becoming embedded in banking infrastructure.
As a result, the institutions deploying it are seeing a genuine and growing competitive advantage. And these faster internal operations translate into better client service, lower costs, and stronger margins. All in all then, the case for continued investment is well established.
But is it all as positive as it seems?
The Applications of AI in Banking Have Limits
When you examine the applications of AI in banking closely, a pattern emerges in that the value is largely internal. Each bank becomes more efficient at processing what it already receives, managing risk within its own systems, and serving clients through its own channels.
The issue comes when you consider how international banking works for the clients relying on it. Because an SME expanding into new markets, a VC-backed business managing operations across several jurisdictions, or a PE-backed company with a complex corporate structure isn’t served by one institution. It’s served by several, each holding a piece of the relationship.
The quality of that client’s experience is therefore shaped by how effectively those institutions interact with each other. And AI investment within any one of them doesn’t change that. Because when Lloyds invests in AI-driven risk management, that investment improves Lloyds’ risk function. It doesn’t improve how risk data is shared, interpreted, or acted upon across a client’s wider banking structure.
A growing business seeking to expand into a new market may interact with three or four institutions through its banking structure. Each will conduct its own assessment using its own models and data inputs. The absence of shared data standards means those assessments remain disconnected. Each institution is working from a partial view of the client, and gaps emerge due to a lack of shared infrastructure to connect the picture.
AI Is Raising Client Expectations Faster Than Coordination Can Keep Up
There’s a secondary consequence that’s easy to miss. As AI becomes embedded in domestic banking, clients are experiencing faster processing, more responsive service, and sharper insights within the institutions they know well. Those experiences set a new baseline for what good banking feels like.
Those same clients then carry that baseline into their cross-border banking relationships. And that’s where the gap shows up. Because the coordination layer between international institutions hasn’t kept pace with what AI is delivering inside individual ones.
So, a client that receives near-instant responses and data-driven analysis domestically will notice when their international banking structure still operates on manual processes, delayed reporting, and disconnected assessments.
In short, AI is raising the bar at exactly the point where fragmented international banking is least equipped to clear it. For mid-tier and regional banks, this is shaping how internationally active clients evaluate their banking relationships.
The Interoperability Gap That AI Investment Can’t Close
As AI capabilities advance across the banking sector, institutions that invest heavily in the technology will become significantly more capable internally. And those that don’t will likely fall behind. The result could be a widening capacity gap between individual banks.
For internationally active clients, this creates an uneven experience across their banking structure. One institution processes in days what another takes weeks to handle, and one provides sophisticated, data-driven client insights while another relies on manual processes.
The issue here is that the weakest link in the structure shapes the overall client experience, regardless of how capable any individual institution is. And this is the interoperability gap. Because real improvements require all of the institutions in a banking structure to operate within a common framework. One that sets a baseline for how data is shared, how processes are aligned, and how client-facing capabilities are coordinated.
Without this framework, AI investment within individual banks could amplify fragmentation rather than reduce it.
What Coordination Infrastructure Adds That AI Can’t Replace
Governed banking networks address a different problem to the one AI is solving. What happens between institutions.
When banks operate within a shared governance framework, the data flowing between them is structured, standardised, and actionable. Client information doesn’t need to be re-gathered or re-interpreted at each institutional boundary. And processes don’t restart from scratch over and over.
For internationally active clients, this changes the nature of the service they receive.
Because, of course, AI accelerates what’s already working. But for mid-tier and regional banks, the strategic question is whether the coordination infrastructure is in place to give AI something meaningful to accelerate.
IBOS Association is a global alliance of independent banks operating across more than 38 markets, delivering coordinated cross-border capabilities through its banking network services. Its shared governance frameworks, aligned data standards, and coordinated banking infrastructure provide the foundation that allows member banks to deliver a consistent, high-quality experience to internationally active clients, across every institution in their banking structure.
And as AI investment continues to improve functions within individual institutions, the banks operating within governed networks are best positioned to translate internal capability into something real for their clients.
Get in touch with Manoj Mistry to find out how IBOS membership gives your institution the coordination infrastructure to make AI investment count across borders.