Banks have seen it all before. When digital communications took hold in the early 2010s, institutions raced to bolt on point solutions: one tool for email, another for SMS, a third for web chat, a fourth for mobile push. Each solved a narrow problem in isolation. A decade later, many banks are still untangling the resulting mess of disconnected systems, inconsistent customer data and integration debt that eats budget faster than it delivers value.
AI is on track to repeat this pattern, and the warning signs are already visible. A typical bank could soon be running as many as 30 AI agents, each operating in silos, with no shared rules for how they interact. Picture a document intake agent, an origination agent and a servicing agent quoting the same borrower different terms, or an unsupervised underwriting agent that's been making poor recommendations for months. This is the direction institutions are heading when AI is procured one point problem at a time.
The ROI gap is a symptom, not a surprise
The financial cost of this approach is already showing up in the numbers. A recent MIT study found 95% of organizations see no measurable return on their generative AI investments, despite banks pouring billions into the technology. Institutions are treating AI as a series of IT projects bolted onto existing workflows rather than as infrastructure that needs to support many workflows over time.
Off the shelf tools solving one narrow use case force banks into an uncomfortable choice: deploy dozens of disconnected point solutions and accept the resulting "agent sprawl," or build custom frameworks demanding constant technical maintenance and rarely translating cleanly from one department to another. Either path tends to generate technical debt faster than it generates value.
Building the capability, not just the fix
The alternative isn't slower AI adoption. It's investing in the underlying capability rather than the immediate use case. This means the collections platform, the fraud model or the customer service agent shouldn't just handle the problem in front of you. It should be architected so the same decisioning engine, data model and governance layer can be extended to the next problem without another rebuild.
This is the distinction between AI that's "AI native," meaning the architecture was designed from the ground up to make AI a core decisioning function, and AI that's merely "AI enabled," where models are layered onto legacy systems as an afterthought. In an AI native system, an improvement to the underlying model or data pipeline improves every process that touches it. In an AI enabled system, each improvement is confined to the module where it was bolted on, and the next use case starts from scratch.
Analysts are converging on a similar prescription for banks navigating this shift. Lenders deploying AI agents need to define a domain of authority for each agent, a deference protocol for when agents should defer to another agent or a human and a resolution layer where conflicts get arbitrated across the portfolio, rather than leaving every agent to operate as its own silo. That framework-first thinking, rather than tool-first thinking, is what separates institutions seeing measurable AI ROI from those stuck rerunning pilots.
Applying the lesson to collections
Collections is a useful proving ground for this principle because it touches risk scoring, customer communication, compliance and payment processing all at once. A bank that buys a standalone chatbot for self service payments and a separate model for delinquency scoring is repeating the same fragmentation that plagued digital channels a decade ago. A bank that invests in a decisioning layer gets a platform that can absorb new use cases, from hardship assessment to recovery optimization, without starting over each time.
The lesson from the digital communications era wasn't that point solutions are bad; it's that infrastructure decisions made under time pressure are hard to unwind. Banks now have a chance to make a different choice with AI: build the capability once and let it compound.
About C&R Software
Trusted by seven of the top 15 U.S. banks, C&R Software is the industry leader in credit risk and collections technology. Debt Manager's AI native solution leverages the latest technology to humanize collections and recovery from end to end. Learn more at https://www.crsoftware.com.
