As Missed Card Payments Rise, FICO Recommends using AI in Collections to Tackle Rising Volumes and Outdated Processes

Analysis of UK 2025 credit card data by global analytics software leader FICO (NYSE:FICO) has thrown the spotlight on the challenges facing collections teams: 

• The percentage of customers missing two and three payments trended upwards from May and June 2025 respectively, for the remainder of the year; balances also increased year-on-year.
• The average balance for accounts with two missed payments peaked at £2,938 in November 2025 (+4.9% YoY) and those with three missed payments rose to £3,324 in December (+4.1% YoY).

With increasing volumes and growing operational pressure, FICO has identified a significant opportunity for artificial intelligence to transform collections strategies, improving efficiency, reducing costs and delivering better customer outcomes.

Many collections operations today remain heavily reliant on manual processes and human intervention, with agents handling cases individually. This approach is difficult to scale, expensive to maintain and often fails to deliver consistent customer experiences, particularly as volumes rise and customer situations become more complex.

“Collections is one of the clearest examples of where traditional processes are no longer fit for purpose,” said Mike Trkay, CIO at FICO. “Lenders are dealing with high volumes, constrained resources and increasing customer expectations. AI offers a way to fundamentally rethink how these interactions are managed.”

While earlier automation efforts such as diallers and basic chatbots have improved productivity, they have often been limited to simple, scripted interactions. These tools typically lack the ability to understand customer context, capture nuance or adapt to individual circumstances, restricting their effectiveness in more complex scenarios.

Trkay points to the growing role of conversational AI as a step forward, enabling more natural, two-way interactions between lenders and customers. By understanding customer intent and capturing key information in real time, conversational AI can streamline processes and reduce the need for manual intervention.

Conversational AI can better understand stress factor clues that a customer may present, as vulnerability isn’t always obvious. For example, someone who normally pays on time may be more stressed when they miss a payment due to affordability pressures than an habitual sloppy payer. They may need to have more specialised collections treatment to navigate them through this challenging time, which also goes a long way to keeping them as a customer in the long term.

There are challenges emerging within conversational AI, however. As it can track with a customer in any direction, there’s a greater risk of ending up with inconsistent (and potentially bad) customer outcomes. It’s critical to properly codify policies and standards within the conversational AI.

AI agents also need to be limited in terms of their language and guidance, and they should recognize when they are on the verge of presenting advice, or their recommendations are becoming too prescriptive. “Basically, the AI needs to know when it should stop and make a hand-off to a live agent,” Trkay says.

Trkay emphasises that the greater opportunity lies in combining conversational capabilities with decision intelligence, allowing AI not only to understand customer needs but to take action and optimise outcomes dynamically.

This includes:

• Enabling automated negotiation of payment plans
• Identifying customers experiencing financial hardship
• Guiding individuals through tailored repayment options
• Ensuring compliance with regulatory requirements and internal policies

“AI enables organisations to move beyond static, one-size-fits-all approaches,” Trkay added. “It allows lenders to deliver more personalised, responsive and effective collections strategies, improving outcomes for both the business and the customer.”

In parallel, AI-driven optimisation techniques are helping lenders make more effective use of limited resources. By evaluating trade-offs between cost, effort and expected outcomes, organisations can prioritise the right actions for each account and deploy capacity more efficiently across large portfolios.

As economic pressures continue and collections volumes fluctuate, FICO warns that lenders that fail to modernise risk falling behind. Those that embrace AI-driven approaches will be better positioned to scale operations, manage risk and deliver more empathetic and effective customer experiences.

“Organisations that leverage AI effectively will not only improve efficiency but also build stronger, more resilient customer relationships,” Trkay concluded.