It’s hard to have a conversation about the future of payments right now without AI coming up.
What can we automate? Where can AI help? What are other financial institutions doing? Are we falling behind if we’re not doing something with it?
They’re fair questions. But there’s another question worth asking first: What problem are we actually trying to solve? From there, we need to consider how the solution fits the institution’s broader goals.
Imagine a payments team spends hours every week reviewing exceptions. The immediate thought might be that AI could make that process faster. But why are there so many exceptions in the first place?
Maybe the system is generating more alerts than the team can reasonably review. Maybe information lives in several places and has to be manually pieced together. Or maybe a process designed years ago has accumulated extra steps along the way.
AI might help. So might changing the process. Either way, the decision should account for what happens beyond that one task.
Sometimes you need to step back before moving forward
At NEACH Payments Group, we help financial institutions look at the bigger picture, whether that’s a payments process, a risk management practice, a BSA/AML program, or a treasury management strategy.
Sometimes the opportunity is new technology. Other times, it’s getting more from what you already have, removing unnecessary manual steps, or strengthening a control. That same thinking should guide the AI conversation.
Understand the process before you automate it
When something takes too much time, it’s natural to look for ways to make it faster. First, ask why it works the way it does.
- If information is being entered multiple times, why?
- If an approval requires several people, what risk is each approval addressing?
- If employees maintain spreadsheets outside the primary system, what do those spreadsheets provide that the system doesn’t?
- If a process has ten steps, are all ten still necessary?
Understanding the workflow helps you identify where technology could improve it and which controls need to remain in place. That matters in payments, where one process can touch operations, risk, compliance, customer experience, staffing, and technology.
Document the why and prepare for disruptions
Those questions also reveal why documentation matters. A separate spreadsheet, duplicate entry, or workaround may address a system limitation, support reconciliation, provide information another team needs, or help employees operate during an outage. It may also be a step that is no longer needed.
Written procedures should capture both how the process works and why key steps are performed. Documenting the purpose of workarounds, duplicated information, and controls helps institutions determine what to retain, improve, or replace as technology changes. It also helps prevent an efficiency improvement in one area from creating a gap elsewhere.
Documenting a workaround does not mean preserving it indefinitely. It creates a basis for evaluating whether it remains necessary and appropriately controlled.
The banking disruptions on November 16, 2026, including reports of transaction delays and some customers being turned away at Chase branches, highlighted why contingency planning belongs in these decisions. As institutions add AI and automation, employees still need to understand the underlying process and know what to do when supporting systems are unavailable.
Procedures should identify which activities can continue through approved alternatives, which must pause, who can authorize changes, and which controls must remain in place. Those instructions need to be accessible during an outage, and employees should practice the alternatives and recovery steps. Not every activity can continue manually, but understanding those limitations in advance can help minimize operational and customer impact.
Make the most of the technology you already have
Financial institutions have already made significant investments in technology for fraud monitoring, BSA/AML, payments, treasury management, reporting, and risk management. Many of these tools connect with the institution’s core system.
Before adding another capability, ask: Are we getting everything we can from what we already have?
At NPG, we help institutions evaluate how their systems, workflows, controls, and business objectives fit together, and where changes can make the greatest difference. Sometimes a new capability is needed. Other times, the opportunity already exists inside a system the institution owns.
Start with a clear outcome
Are you trying to detect fraud sooner, reduce unnecessary manual reviews, improve how risk is communicated, simplify treasury onboarding, or give employees better information for decisions?
Once the outcome is clear, you can evaluate whether AI, other automation, existing technology, a process change, or a combination of these approaches makes sense. Define what success looks like, such as fewer exceptions, shorter onboarding times, or less rework, and assess the effects on other teams, controls, and customer service.
Consider how the solution fits the bigger picture
Solving a specific problem is a useful starting point. But a technology investment should also support the institution’s broader direction. A tool that saves time for one team may create extra work for another, introduce a new dependency, or add another place where information must be maintained.
For example, AI might speed up the review of treasury onboarding documents. But if employees still have to reenter the results into several systems, or downstream teams cannot use the information, the overall process may see little improvement. Evaluate the full workflow, from the customer’s initial request through approvals, setup, and ongoing servicing.
Will the capability work with existing systems? Support other business priorities? Improve the experience across departments? Can the institution manage its costs, controls, staffing needs, and contingency arrangements as its use grows?
A capability does not need to solve every problem to provide value. It should address the immediate need in a way that fits the broader operating model and moves the institution toward its goals.
The question becomes: Does this solve the problem in a way that makes the institution stronger overall?
AI should support the strategy
Financial institutions should be exploring what AI could mean for their organizations. The objectives should guide those decisions: reducing risk, improving processes, strengthening operational resilience, and delivering a better customer experience.
Before moving forward, ask:
- What would we like to make better?
- How does the solution fit our broader strategy and operations?
- How will we manage the process if the technology is unavailable?
Together, those questions help institutions choose capabilities that deliver value across the process, fit their long-term direction, and remain supported by people, controls, and written procedures.
Start the conversation
If there’s a process, workflow, or operational challenge you’re trying to improve, NPG can help you understand the problem, evaluate how potential solutions fit your broader operations and strategy, and determine what makes sense for your institution.