The end of digital banking? Welcome to the cognitive banking era

The evolution: From DIY to AI

The transition from digital banking to cognitive banking can be thought of as an evolution through three stages:

1. Do-it-yourself: Members handle tasks employees once did, navigating digital platforms on their own.
2. Do-it-with-me: AI-powered co-pilots assist members, making self-service more intuitive.
3. Do-it-for-me: AI anticipates needs and takes action proactively.

This is the future of credit unions: moving from reactive self-service to hyper-personalized proactive service.

Digital banking's core problem: Making customers do the work

Digital banking was meant to improve access, but instead, it shifted the work to customers. Institutions built interfaces that let members DIY tasks once handled by trained employees—without rethinking the process itself.

What’s worse, those employees were specialists, focused solely on delivering these services efficiently. In our attempt to enhance the customer experience, we ended up focusing on pixels instead of process. We gave customers screens to complete tasks but left them to navigate complex workflows on their own.

If you’re like me, your expectations have shifted. I use browser-based search engines less and AI-driven chat more. Instead of navigating through a menu, I type a question and get an answer. That’s what members now expect from their credit union.

Do-it-with-me: The AI co-pilot model

Generative AI provides an entirely new approach to digital banking—one that combines expert assistance with 24/7 convenience. Instead of static menus and search bars, AI-driven co-pilots engage members in natural, intuitive conversations.

Early adopters are embracing this co-pilot model, where AI assists both employees and members in real-time. Over time, as trust in AI grows, institutions can transition to full self-service AI agents.

Do-it-for-me: The cognitive banking model

The real breakthrough is predictive service. Instead of waiting for members to take action, credit unions can use AI and data to anticipate their needs and proactively assist.

Take a simple example: updating a beneficiary. I recently discovered one of my investment accounts didn’t list my wife as a beneficiary, while others did. It took me far too long to find where to update it, and I eventually had to call support. 

Imagine a smarter system. It recognizes the inconsistency and sends a secure message:

"We noticed one of your accounts doesn’t have a beneficiary listed. Would you like to update it?"

With a simple “Yes” or “No”, the task is done.

This shift—from Pull/React (customer-initiated service) to Predict/Push (institution-initiated service)—drives efficiency and dramatically improves the member experience.

Overcoming the "we're not ready for AI" mindset

The biggest barrier to AI isn’t technology. It’s hesitation.

Many credit unions assume they need perfect data or a major system overhaul before they can start. That’s not true. AI doesn’t require years of cleanup or expensive upgrades to make an impact. Most credit unions already have the data they need—it’s just locked in core banking systems and scattered across platforms. The real challenge isn’t collecting data, it’s making it accessible and actionable.

This is where AI changes the game. Instead of following the traditional digital transformation playbook, credit unions have a chance to leap ahead. AI can enhance member experiences, automate operations, and improve decision-making without waiting for a full technology refresh.

Platforms like Agentforce, built on Salesforce, make this even easier. If your core system is connected to Salesforce, you already have the foundation to get started. Even with a limited dataset, AI can deliver results—streamlining account servicing, automating routine inquiries, and improving fraud detection.

Regulators are supportive. The National Credit Union Administration (NCUA) is pushing for AI-driven efficiencies, signaling that the shift isn’t optional. The institutions that start now will lead. The ones that wait will be playing catch-up.

“My priorities as Chairman include…Promoting the appropriate use of artificial intelligence (AI) as a tool for NCUA employees. One goal is enhancing productivity, but it’s also true that regulators who use technologies are more apt to understand why the regulated use them.”