How AI is quietly transforming the Payments landscape

The Payments ecosystem is going through a period of fast-paced innovation.

We see a shift from card-based payments to account-to-account models. Embedded finance continues to grow, moving control of both money and customer relationships towards non-traditional players such as super-apps and retailers.

At the same time, the push for ‘instant everything’, combined with ongoing digitisation of currencies, wallets, identities and deposits, is reshaping how Payments operates. And all of this depends upon platforms that can deliver better, faster and more flexible customer experiences. As a result, long-overdue infrastructure upgrades are now high on the agenda.

Against this backdrop, AI is mentioned less often as a primary driver of innovation in Payments than in other financial services domains. That may be temporary. It may also reflect how we define ‘Payments’.

When we look at how financial institutions are scaling AI today, there is a consistent focus on risk and operational efficiency.

Where AI is already delivering value

• KYC document handling: one of the most established use cases. AI helps process documentation more efficiently and supports the shift from onboarding-only KYC towards continuous, lifecycle-based, KYC.
• Real-time transaction monitoring: AI improves the detection of known fraud patterns and helps identify new, previously unseen behaviours. Many institutions are actively testing and scaling these capabilities.
• Investigation triage and prioritisation: AI supports investigators by directing attention to the most relevant cases. It also assists with investigation handling, allowing human experts to focus on higher-value analysis.

These use cases are already in production but often remain out-of-sight. Because they focus on Operations and Risk, they are not always perceived as “transformational” in the same way as customer-facing innovations. Yet their impact on both institutions and customers can be significant.

The same applies to the use of Generative AI in software development. From documenting and reviewing code to supporting both forward and reverse engineering, GenAI is improving speed and quality across IT teams. As these capabilities mature, they could also accelerate large-scale modernisation programmes, including the migration away from legacy COBOL platforms. The business impact of such a transition would be substantial.

Understanding AI as a toolbox, not a single solution

Examples across AI types

Before looking ahead, it helps to clarify what we mean by AI in practice.

• Discriminative AI: used in AML and fraud monitoring, with machine learning models trained on internal and, in some cases, shared datasets.
• Generative AI: supporting investigation-handling by collecting and structuring data, preparing it for human review.
• AI agents: enabling risk-based triage and prioritisation, and allocation of payment investigations to the most suitable expert.
• Agentic AI: emerging use cases include automating KYC data collection and analysis, with suggested decisions for low-risk clients.

Where AI in Payments is heading

The current strong focus on KYC, AML and fraud is likely to remain. At the same time, we expect AI adoption to expand further into operational areas. Key opportunities include:

• Automated reconciliation
• Smart and adaptive payment routing, particularly in cross-border scenarios
• Chargeback management

Beyond this, we see a broader shift taking shape. AI in Payments is moving from behind-the-scenes efficiency gains towards more visible, customer-facing and revenue-driven use cases.

Moving closer to the customer

Today, chatbots and virtual assistants are among the few widely-adopted customer-facing applications. That will change. We expect to see more advanced use cases emerge across the Payments ecosystem:

• Hyper-personalisation: AI will tailor user experiences and product recommendations, from payment methods to card options. It will also optimise authentication, reducing friction for expected behaviour while increasing security when patterns deviate.
• Agentic e-commerce: Personal shopping agents will be able to initiate and execute payments in a controlled, transparent and auditable way.
• Embedded finance: Lending and other financial services will increasingly be integrated into non-financial platforms, apps and marketplaces.

From product to customer: a shift in perspective

Another important shift is from product-centric thinking to customer-centric thinking.

AI makes it possible to build richer, more dynamic customer profiles based on payment behaviour. These micro-profiles can support:

• Customer lifetime value management
• Early detection of churn risk
• Increased use of transactional products
• More relevant next-best offers

This creates opportunities to combine growth and the customer experience in a much more targeted way.

A moment of acceleration and a shift in focus

AI in Payments may have started more quietly than in other domains, but momentum is clearly building.

We now see tangible progress across Risk Management, Operations and customer-facing areas. AI is becoming a driver of revenue growth, stronger risk control, and more efficient and flexible operations.

For financial institutions, the question is no longer whether to use AI, but where to focus next. Those who build on today’s foundations while exploring more customer-facing use cases will be best positioned to capture value as adoption accelerates.

For us, this evolution reflects both the pace of change and the growing level of innovation across the global Payments landscape.

Wherever your organisation is on that journey, the question is the same: where should you focus next? If you are exploring how to move beyond pilots and scale AI in a way that works for you, we are happy to share our experience and discuss what this could look like in your specific context.