Banking's AI Revolution Starts in Operations?

Written by Hocine Merrir, Associate Partner, Lead AI in Banking Operations, and David Steiger, Associate Partner, Head of Next Gen Operating Model

While public attention focuses on copilots and chatbots, a quieter, larger opportunity is taking shape where few people look. A new Synpulse study, AI in Banking Operations, surveyed 26 Swiss and Liechtenstein banks and service providers and found that AI's potential in operational banking processes may outweigh the front-office spotlight.

Across six core operations domains and 25 activities, 78 percent of operational tasks are now considered relevant for AI, and 43 percent of responses point to high or very high expected benefits. Given the volume and complexity of work running through banking operations, a number this high signals more than curiosity – it signals operations could become a primary arena where AI reshapes how banks function.
Deployment is still catching up.

The study identified 37 concrete AI initiatives across the surveyed institutions, including 11 already in production, with two-thirds of institutions having started at least one. The gap between interest and deployment is real – but it looks like an early-moving market, not a disinterested one.

Where the Opportunity Is Most Concrete

Some activities already show what this shift could look like at scale. Corporate action processing combines the highest AI relevance score in the study (96 percent) with the strongest expected benefits and the most advanced production deployments observed. Teams that once manually interpreted event notices, tracked elections and prepared client communications are beginning to supervise AI systems handling this work directly. Reconciliation, payment processing, credit origination and asset transfer processing follow closely behind.

The pattern is consistent: AI's potential is highest not in repetitive tasks, but in those requiring interpretation and judgment. Document-heavy, exception-driven processes are where the opportunity is most concrete – and where the first proof points are emerging.

The shift is also being shaped from outside individual banks. Service providers, fintechs and platform vendors are embedding AI directly into infrastructure many banks already rely on. For smaller institutions, AI-native capabilities may arrive through the ecosystem whether or not they run a single internal pilot.

From Potential to Practice

Adoption is selective today, with most institutions still in early scaling stages, struggling to close an ever-widening gap between AI ambition and operational adoption. The report's authors argue the next phase is not about generating more ideas – the use cases are already identified – but about deliberate choices: which processes to redesign around AI, how to build data and governance foundations that make scaling durable, and how to engage vendors and platforms as active partners rather than passive suppliers.

The scale of the opportunity is itself a reason not to wait. As the report puts it, banking operations have for decades run on one principle: humans decide, systems execute. AI is starting to invert that – with systems handling first-pass interpretation and humans supervising exceptions. Given how much critical workload sits in banking operations, that inversion would be one of the more consequential shifts in how banks run.

The Opportunity Is Open – For Now

Swiss and Liechtenstein banking is still early in this shift compared to global leaders, and investment levels reflect a cautious, targeted approach rather than broad transformation. But the scale of the opportunity, the clarity of the use cases and the first proof points already visible suggest more than a passing trend. Institutions building data foundations and scaling proven use cases now are positioning themselves to lead as AI in operations moves from potential to standard practice.

The full study is available exclusively to survey partners. Contact Synpulse to request access to the key findings or to discuss how AI can transform your banking operations.

Contacts: Hocine Merrir, Associate Partner, Lead AI in Banking Operations, and David Steiger, Associate Partner, Head of Next Gen Operating Model