Now or never: Turning AI into Competitive Advantage by TRENT GILLESPIE
FOR NATIONAL ASSOCIATION OF CREDIT UNION SERVICE ORGANIZATIONS
Thanks for attending my keynote talk!
Here are the additional resources I promised:
- Key Slides
- My free Generative AI at Work AI Course.
- My Future Customer Analyzer to identify your Future Customer and how they change.
- My AI Strategy Assessment & 90 Day Action Plan tool
- My AI Roadmap Creation Tool for CUSOs.
- Below is the AI summary of the keynote talk.
Need help with AI? Here is a link to the AI Jumpstart service through my company, Stellis AI.
Have other questions? Contact Trent
Now or Never: Transform to
Serve Your Future Member
The Moment You're In
Every leader in the room has a conference room moment — a point where the opportunity is bigger than your comfort level, and you have to choose: wait until you're ready, or lead anyway.
AI is that moment. Right now.
The CUSOs and credit unions winning with AI are not the best-funded or most technical. They're the ones experimenting, learning, and moving — ready or not. And what they're not telling you is they're having fun doing it, solving problems that have been stuck for years and creating products and services that never felt possible before.
Two Mistakes Most Leaders Are Making
Mistake 1: Treating AI as a tool rollout
AI is not a CRM upgrade or a new dashboard. It is a complete rewiring of how your business operates — your products, services, cost structures, required skills, and potentially who your customer is. Tool rollouts don't require leadership. This does.
When organizations combine a culture of innovation with AI to scale it, they unlock not just productivity but the ability to give their people power to become builders of the future, not bystanders. That combination is what drove Amazon's growth, and due to generative AI, it no longer requires billions of dollars. It's $30 a month. But it does require three things at the core of Operational AI™: leadership, vision, and action.
Mistake 2: Optimizing for a world that no longer exists
Most leaders are applying AI to how they worked five, ten, or twenty years ago instead of reinventing for how they need to serve customers in the future. When members are expecting AI-enabled financial services as a primary channel, vision is no longer a luxury. It is the foundation of reinvention.
Where the Industry Stands Today
| Signal | Data Point |
|---|---|
| Daily AI users reporting productivity gain | ~20% improvement in productivity and quality of work |
| Weekly AI users | ~5% improvement — almost negligible |
| Less than weekly | No measurable benefit at all |
| Knowledge workers now using AI on the job | ~50%, even without official authorization |
| Consumers using AI for financial planning/budgeting | 55% |
| Of GenAI users, Gen Z and younger using AI for financial decisions | 82% |
Daily use is the first competitive edge. And at 50% adoption, non-users are now the minority.
Three things to take away: people are finding real value (that's why they're going against company policies to use it), the tipping point has arrived in just the last month or two, and if your organization is not empowering employees to use these tools, you're leaving opportunity on the table and absorbing risk if people don't handle data appropriately.
The key message: Get AI out of the shadows and empower it. But just giving your employees AI is not going to reinvent your company — it won't change your products, services, or how you compete. For that, you need vision.
The Core Question: Your Future Customer™
At Amazon, vision was created through one relentless question: what will our future customer want five years from now? Then work backward from it to build it before anybody else. That's how Amazon disrupted industries and invented Alexa, the Kindle, and reinvented e-commerce.
Your members' expectations are already changing. In five years, younger generations will have been using AI for eight years, potentially the majority of their adult lives. They're going to expect AI-enabled services everywhere.
As CUSOs, you face a dual challenge: use AI effectively in your own organizations AND empower your credit unions to serve members who will have these expectations. That's what makes your role so critical.
The Future Customer™ Method: Five Questions for Your Leadership Team
- What do your customers say they value most today? Not what you think — what they tell you.
- Which of those values will still differentiate you when know-how is free and instant? If expertise, speed, decision-making, and personalization are now done by software, what do your customers still value you for?
- Which of those values can you no longer charge extra for — because it becomes table stakes?
- Where will customers still desire and pay for a uniquely human touch?
- What brand-new value could you create because intelligence is cheap — something customers don't even ask for yet?
The first three questions establish your baseline and expose what's at risk. Questions four and five are about building for the future — they become the foundation of the Trailblaze step in the AI SPRINT™ framework.
Example shifts already forming for CUSOs and the credit unions they serve
| Past Customer Expectation | AI-Powered Shift | Future Customer Expectation |
|---|---|---|
| Shared services and cost reduction | Intelligent agent automation | Eliminate the work, not just the costs |
| Operational support | Predictive intelligence | Manage my risk in real time with perfect visibility |
| Member growth and retention | Hyper-personalization at scale | An AI for every single member, trained on their history and every product the credit union offers |
What you need to do is start building your organizational vision for those future expectations — not rebuild the past, because those old expectations are gone.
What Changes When Intelligence Is Cheap?
For the entire history of the human race, intelligence has been the most expensive resource organizations buy. Leaders are hired and compensated well. Organizational structures are built around routing decisions through people, hoping for intelligent outcomes. And there's never enough.
Now it's cheap, instant, and on demand — $30 a month. Think of it like electricity, but for thinking. Except electricity waits for you to plug something in. AI is actively watching, monitoring, and taking action.
Live Demonstration
Trent demonstrated building a client account risk monitoring tool — designed to identify accounts at risk of leaving so the organization can preserve revenue. He prompted ChatGPT to design the application (generating what used to be called business requirements), then pasted those into Replit, an AI app builder, and said "build this."
The result: a working dashboard showing 50 client accounts scored by risk level (23 healthy, 15 at risk, 12 critical), with drill-down detail on each account, the specific reasons behind each score, and a live outreach campaign builder that generated personalized email, phone, and in-person scripts customized to each client's situation — including flagging that one account had 82 recent support tickets as a likely driver of churn risk.
The question posed to the room: how many months would it take your organization to decide to build a tool like this? When any employee can design and implement a process improvement in 12 minutes for less than the cost of Starbucks, it's not a productivity change. It is a reinvention of work. The only question is who does it first — you, your employees, or your competitors.
Reinventing Processes with AI at the Center
The Amazon Tokyo Example
Trent described automating last-mile delivery in Tokyo, where drivers were plotting routes on paper maps. His team trained AI on Tokyo's address scheme, package sorting logic, vehicle routing, and time-of-day traffic patterns. Drivers showed up to work with an app where AI had already optimized their entire day.
Everyone benefited: drivers went home on time, customers got reliable deliveries, Amazon reduced costs, and the reinvention enabled tens of thousands of new jobs in Japan. The lesson: AI-driven reinvention can scale beyond human ability while supporting all stakeholders — but only when processes are redesigned from the ground up.
The Agent Landscape: Two Types to Know
General Purpose AI (ChatGPT, Claude, Copilot, Gemini)
Think Swiss Army knife. These are becoming the primary interface to the entire workday, potentially replacing email, ERP, and CRM systems. Even Salesforce recently announced they will no longer require a user interface to access their system at all — AI agents will do it. These tools give individuals superpowers, but the problem is that 10 people using general-purpose AI on the same task will produce 10 different results.
Specialty Agents (purpose-built for your organization)
Think power tools. Designed for specific purposes, trained on your data, and deployed like a team member. One organization-standard agent produces consistent results across every person, every time.
High-value agents for CUSOs and credit unions:
- Member acquisition agent — responds to new member inquiries and potentially handles applications
- Portfolio management agent — flags members with early signs of distress to intervene before problems develop
- Board meeting agent — automates meeting preparation and document assembly
- Regulatory monitoring agent — tracks regulation changes and helps manage compliance work to avoid findings
These aren't a far-off future. Anyone in ChatGPT or Claude can create agents today. Trent demonstrated an agent autonomously logging into HubSpot, navigating the interface by sight, finding a contact, creating a follow-up task, and handling a date-picker — end to end with no human input. ChatGPT now allows any employee to create an agent and share it across their organization.
Trent also shared a list of AI agent use cases available to NACUSO members for less than $500 per month, including FIT scoring, proposal intelligence, relationship signal monitoring, regulatory checklists, and staff onboarding agents.
Reinventing Processes with Agent Chains
The real power is not individual agents but chains of specialized agents producing consistent, repeatable outcomes.
Sales reinvention example: A prospecting agent works 24/7 scanning for signals of potential prospects. When it finds one, a qualifier agent evaluates budget, service area, and profile fit. Qualified prospects pass to a research agent that pulls public earnings, LinkedIn profiles, and every available data point to create a hyper-personalized sales strategy and script. Salespeople arrive at work with a list of qualified prospects and a tailored approach for each one. This changes sales entirely — conversations are better than they've ever been possible before.
Commercial lending intake example: A document collection agent chases down and validates borrower materials. A credit file agent flags exceptions and assembles the package. A status communication agent keeps all stakeholders informed. Humans handle the higher-level judgment calls.
Scale, Swarms, and the Agent Economy
Agent Swarms
Everyone is going to be a manager of AI agents — not one or five or ten, but potentially hundreds. AI agents can hire other AI agents (swarms), and there is even a startup allowing AIs to hire humans when AI can't do the job.
Agent-to-Agent Commerce
In the future, your customers may never view your website, email, or call you again. Their AI agent is going to decide if you make the cut.
Trent walked through a specific CUSO example: a credit union tells its AI, "Find commercial lending CUSOs serving credit unions under a billion dollars with SBA 7A capability at a maximum 45-day close." The AI finds a CUSO's agent, which responds with a 38-day close and 6.75% participation rate and offers to start an application. The transaction proceeds through agents. There is literally a bank — Meow AI — that allows AI agents to sign up for checking accounts and credit lines right now.
This is starting with B2C use cases, with B2B expected in the next 12 to 18 months. In the future, you won't just service humans but their AI agents too. If you're not at least thinking about this and preparing, you're probably not preparing for the future of commerce.
What's Left for Humans? Vision.
When intelligence isn't the limit anymore, imagination is. The ability to see opportunities, create the plan, and bring the entire organization along — not just at the CEO level, but at every level. The edge will come from people closest to the work who see customer problems and respond faster than anyone else.
When the world's biggest financial services companies are moving like this, your credit unions will be at a disadvantage if you do not enable them. That is your opportunity — to provide through shared infrastructure the same types of capabilities so credit unions can continue to compete and support their local communities.
The smartest people in the room won't be those who know the most but those who know how to think and innovate with AI.
The AI SPRINT™ Framework
The AI Sprint is not a project — it's a rhythm, a discipline of continuous learning, experimentation, and sharing. Every month: what is our best opportunity to use AI? Try it. Reflect on what you learned. Adjust, share, and do it again. April, May, June — continually accelerating. This isn't just for leaders. Everyone can do an AI Sprint. It's as simple as learning and sharing.
But speed without direction is a treadmill to nowhere. That's why you need vision first.
Light the fire that gets AI moving. At Amazon, the biggest problems were often solved not by the people with the biggest titles but by one person who became the spark.
- Educate leadership together. Hands-on training as a team, not a keynote or online course they won't complete. When leaders learn together, they start bouncing ideas off each other and go from hearing about AI to leading with it.
- Assign a transformation owner. The AI champion — often someone outside IT, because this spans hiring, incentives, products, and services across the whole organization.
- Start transformation sprints. The monthly expectation of experimentation, learning, and sharing.
Ask yourself: who is your spark? If nobody, it may need to be you.
Design your company for your future AI-enabled customers. The companies winning aren't asking how to save 5% — they're asking how to win the next phase of their industry.
This is the step where you answer the first three Future Customer™ questions, set a timeline, assign owners, and start working. One client set a goal of 10% additional revenue from AI-enabled services within three years, identified over 120 ideas, narrowed to six, invested in two, and those products are launching in the coming months.
Great companies don't optimize for today. They position for what's next.
If Spark lights the fire and Position sets the direction, Rally creates momentum. You can't force teams into the future or create innovation through fear. But you can create an environment where AI use is safe, supported, expected, and rewarded. A recent survey found 69% of employees are scared to tell others about their AI use. That kills innovation. This step fixes it.
Make it safe:- Use business-grade AI tools under business terms that protect your data.
- Create a one-to-two-page AI acceptable-use policy: approved tools, what employees can and can't do with company data. Keep it simple.
- Address job safety directly. Orient employees toward growth: reinvest savings into new products and services. Share the concept of demand creation — AI lets you serve customers who were previously unprofitable, expanding total demand.
- Provide role-based training with 5 to 10 specific AI practices per role, grounded in their actual work.
- Direct communication from leaders that AI use is expected. Consider growth-oriented AI questions in performance reviews: How often did you integrate AI? What was better because of AI? Did you share learnings? Did you experiment?
- Spot bonuses, gift cards, or annual incentives to push past the adoption plateau.
On AI's environmental footprint: One AI message consumes about 1/15th of a teaspoon of water and 0.34 kilowatt-hours of electricity. For 500 moderate daily users over one month, that's 347 gallons of water and 340 kilowatt-hours — less than one person's daily water consumption and less than charging a Tesla six times. Share this with employees to address concerns.
AI can apply to almost any task, and there are over 45,000 AI tools on the market. Start simple — revenue generation first:
- Marketing content and personalization. Organizations using AI in marketing are seeing roughly 400% productivity improvement.
- Sales personalization and research. Not more spam — hyper-personalized sales. One Fortune 500 salesperson doubled their productivity and sales goals using the research agent approach.
- Operational bottlenecks. Apply AI where it matters most. Create a prioritized list and work down it.
You don't need massive investments to start. You don't need a million-dollar data lake. Start with tactical processes while you work on bigger efforts.
Amazon was successful because it created a culture of innovation and used AI to scale it. AI gives employees a 100x increase in their innovation capability. Combine that with the AI Sprint rhythm, and you create a culture of innovation for the first time.
- Establish the sprint rhythm.
- Create an idea pipeline — a place for people to submit ideas.
- Reward the results.
The AI Sprint Email: Once a month, managers ask employees for a five-minute email:
- What did you use AI for this month?
- What changed because of it?
- Who did you share it with?
- What's one task you'll test with AI next month?
This creates accountability and documents what everyone is learning so it can be shared broadly.
Now invest the savings and new capacity into products and services that weren't possible before. This is where Future Customer™ questions four and five become your guide:
- Where will customers still desire and pay for a uniquely human touch? Human connection will be more important than ever. Build AI to support it.
- What brand-new value could you create because intelligence is cheap — something customers don't even ask for yet? This is the question Amazon focused on relentlessly. It's what enables invention, not just optimization.
This isn't just good strategy — it's fun work. Wouldn't you rather spend your time inventing the future than automating old processes?
Action Items
- Conduct hands-on AI training as a leadership team and appoint an AI transformation owner.
- Publish a one-to-two-page AI acceptable-use policy and designate approved business-grade AI tools.
- Build role-based AI training with 5 to 10 specific AI practices per role.
- Apply AI first to marketing personalization and sales research for immediate revenue impact.
- Launch the monthly AI Sprint email across all teams.
- Answer the five Future Customer™ questions and set a measurable AI-driven goal with a timeline and owner.
The Choice
More than 50 million Americans lack full access to financial services. Credit unions were built on the idea that together, members can do what none could do alone. CUSOs carry that same mission forward. Now you have the best tools humans have ever invented to extend that reach further than ever before, to those who have not been able to access your services.
You may not feel prepared. But that's not the question. The question is: will you lead to the best of your ability and do it in a way that creates growth and opportunity for as many people as possible?
At Amazon, every day is Day One — starting with the hunger to make a difference, the passion to serve customers better, and the courage to take on a little bit of risk. Today is your Day One: not to use AI to write an email, but to spark new ideas, invent new products and services, reinvent your industry, and strengthen the communities your credit unions serve.