The opinions expressed here are those of the authors. They do not necessarily reflect the views or positions of UK Finance or its members.

Generative and traditional AI are no longer just bolt-on features - they are becoming essential elements in routing, risk, liquidity, and the customer experience. Stronger authentication at the frontend, real-time fraud interdiction in the middle, and self-tuning liquidity management at the backend, all draw on continuous-learning systems. And banks that do not incorporate this in to their operating model are going to face the same margin pressure and speed problems that mobile-first fintechs created for the Payments sector ten years ago.

AI is forcing a redesign of the Payments landscape. AI models have improved fraud detection by catching suspicious activity in real time - although the model is far from perfect, with global online payment fraud hitting £33bn in 2024 (of which £2.5bn was in the UK) – by offering new ways of authorising payments (like using biometrics, face-pay and voice-activation), and by helping institutions to streamline the end-to-end process itself. Perhaps most impactful of all, incorporating AI is reducing processing errors, with employees spending less time reviewing and approving payments and more time delivering better service to their customers.

Most large institutions still today process high-volume flows on monolithic stacks designed for batch processing rather than for real-time events. Exception queues, low straight-through processing rates and opaque status messages all remain real and present despite years of supposed ‘digital transformation’. AI-driven operating models require deep rewiring - Payments operations demand observability of every event, Risk and Compliance systems need to consume those events in real time, Treasury and Liquidity engines must predict funding needs seconds ahead and data platforms need to broadcast ISO 20022-rich messages tagged with Unique End-to-end Transaction References (UETRs) instantly. Without this deep rewiring, AI can only operate at the edges of the Payments operating model.

How Is The Payments Industry Responding?

Supervisory frameworks from the ECB and the PRA, and the Digital Operational Resilience Act (DORA), are requiring auditable controls around every production model. Institutions that want to scale safely should therefore implement cross-functional AI councils (spanning Risk, Compliance, Operations and Technology) and maintain single inventories that auto-record every model release through Continuous Integration/Continuous Deployment (CI/CD) pipelines. They should also use time-boxed approval cycles - so governance does not become an organisational parking brake - and embed human-override ‘kill switches’ into workflow tools rather than simply burying them in source code.

And How Should Your Institution Respond?

Those institutions currently considering their next Payments operating model steps should choose to be bold and brave in relation to AI, and could do worse than applying the following practical steps:

1. Build an AI model and implementation plan that works now and can evolve
2. Pick the right tool(s) for the job
3. Plan for regular updates to your AI model
4. Enable progress rather than just try to control it
5. Understand and benchmark against what others are doing
6. Measure those metrics that matter to your leadership
7. Update your systems thoughtfully rather than urgently

What will differentiate the real sector leaders will be treating AI as infrastructure and not as experimentation. We today see institutions like JPMorgan, HSBC and Revolut already pushing hard in this direction and publicly sharing the benefits generated as a result. This proves that an AI-driven Payments operating model is not a theoretical proposition – each required component of it already exists in production somewhere. And so those treating this as just another ‘software upgrade’ will simply fall behind their more nimbler and visionary Payments sector competitors.

At Projective Group, we thrive on leading change in the financial sector. To read Alan Verschoyle-King (U.K. Payments Practice Lead for Projective Group) and Stephen Peters’ (Head of Enterprise Payment Solutions at FIS) extensive article on the future of payments, download the Journal of Financial Services.