Think it’s hard to convince new members to open an account at your credit union?

Think again.

Do what the big FIs – Chase, Wells Fargo, Citibank – routinely do: offer cash incentives for joining the institution.

How much?  $400 to $500 are numbers often seen and sometimes the bounty climbs to as much as $1000.

Are credit unions in this too? Indeed they are, and on the show is Fred Cadena, a senior vice president at Vericast, a company that check printer Harland Clarke transformed into  Upwards of 4000 banks and credit unions are Vericast customers and a big service they offer involves strategies for attracting new members and customers.

And money definitely talks in this process.

But a key is offering it intelligently.  Cadena explains that this typically involves tasking the new member with doing a task or two such as setting up a recurring automatic deposit – a pay check is ideal – or perhaps swiping a debit card X times.

Financial bonuses can also help sell CDs, loans, saving accounts – there are lots of uses for these bonuses, Cadena elaborates.

The post CU 2.0 Podcast Episode 426 Vericast Tells How to Supercharge Acquisition of New Account Holders appeared first on CU 2.0.

Have you ever given an AI tool a vague prompt and gotten a vague answer back? Or pasted in one carefully written example and watched the output suddenly become exactly what you needed? That's not luck. That's the difference between zero-shot, one-shot, and few-shot prompting. Once you understand it, you'll get better results from AI every single time—no new tools required.

What Do These Terms Actually Mean?

These three phrases describe how many examples you give an AI model before asking it to do something. That's it. The jargon sounds complicated. The concept is not.

Zero-shot means you give the AI no examples at all. You ask it to do something and trust it to figure out the format and tone on its own.

One-shot means you give it one example of what good output looks like before making your request.

Few-shot means you give it two to five examples. More than that and you're usually into diminishing returns.

Think of it this way. If you hired a new loan operations analyst and said, "Write a member denial letter," you'd probably get something generic. But if you handed them two letters you'd already approved and said, "Write one like these," you'd get something you could actually use. AI works the same way.

The quality of your AI output often has less to do with the tool you're using and more to do with how much context you give it. One good example can do more than a paragraph of instructions.

When Zero-Shot Is Enough

Zero-shot prompting works well when format and tone don't matter much, or when the AI already has strong training on the topic. Ask Claude or ChatGPT (as of 2025—verify current capabilities) to summarize a vendor proposal you've pasted in, and you'll probably get a decent summary without supplying any examples. The task is clear. The format is obvious. Zero-shot is fine.

Other places zero-shot tends to work at credit unions:

• Drafting a first-pass agenda for an all-staff meeting
• Translating a dense policy document into plain-language bullet points
• Getting a quick answer to a general question (always verify compliance-related responses with your compliance team before acting)

When stakes are low and format is flexible, don't overthink it. Just ask.

When One-Shot Pays Off

One-shot prompting is where most credit union executives start getting real lift from AI. You're giving the model a single anchor point—something concrete to pattern-match against.

Say your marketing team wants to draft a new member onboarding email. Instead of writing a long prompt about brand voice, tone, and what not to say, paste in one email your team already loves. Then say: "Write a new version of this for our home equity loan campaign." The AI now has something to work from. The output will be closer to what you actually want.

One-shot is especially useful when:

• You have an existing template—a board memo format, a standard member communication, a loan approval letter
• You want to match a specific voice or reading level
• You're filling in something formulaic but brand-sensitive

How to Use Few-Shot Prompting for Credit Union Work

Few-shot is the heavy hitter. When you need highly consistent output—or when zero-shot and one-shot keep missing the mark—adding two to five examples usually closes the gap.

Here's a practical workflow your team can follow today:

1. Identify the output you want. Be specific. "A two-paragraph member email announcing our new savings rate" is better than "a member email."

2. Pull two or three examples of past outputs you liked. These could be emails, board summaries, exam response letters, or member-facing social posts. Dig into your sent folder or shared drives.

3. Paste your examples into the prompt first. Label them clearly: "Example 1:", "Example 2:", and so on. Don't make the AI guess what they are.

4. State your actual request after the examples. Something like: "Using the style and format of the examples above, write a new version for [specific situation]."

5. Review and refine. AI output is a first draft, not a final draft. Read it before it goes anywhere near your members, your board, or your examiners.

Few-shot prompting is particularly powerful for your member services team when they're handling high-volume, templated communications—rate change notices, overdraft alerts, loan status updates. Build one good set of examples once. Use it again and again.

Why This Matters More Than Which AI Tool You Use

Many credit union executives spend a lot of energy debating which AI platform to adopt. Claude vs. ChatGPT vs. Microsoft Copilot. That debate has its place. But the return on learning to prompt well is often higher than the return on switching tools.

A well-crafted few-shot prompt on a free AI tool will frequently outperform a lazy zero-shot prompt on the most expensive enterprise platform. The skill transfers, too. Whatever tool your credit union lands on, the underlying logic is the same.

Your prompting skills compound. Every good example set you build today becomes a reusable asset for your team tomorrow.

This is especially true for smaller credit unions where one person is wearing multiple hats. A small library of few-shot prompt templates—one for board memos, one for member communications, one for examiner-ready summaries—can save hours every week without adding headcount.

A Quick Reference: Choosing the Right Approach

Not every task needs the same level of setup. Here's a simple way to think about it:

• Zero-shot: You need a quick answer or rough draft, and format flexibility is fine.
• One-shot: You have a template or approved example handy and want the AI to stay close to it.
• Few-shot: You need consistency across many outputs, the task is nuanced, or earlier attempts weren't landing.

Start with zero-shot. If the output misses, add one example. If it still misses, add two or three more. Most of the time, you won't need to go beyond three before the AI clicks into the right pattern. The best prompts aren't necessarily the longest ones. They're the ones that give the AI exactly as much context as it needs—no more, no less.

What's the risk?

The zero-shot, one-shot, and few-shot framework is a prompting technique, not a data integration. On its own, it carries relatively low risk. The bigger question is what you're pasting into the prompt. If your few-shot examples contain member names, account numbers, or any other nonpublic personal information, you're potentially exposing NPI to a third-party AI service. Before you build example libraries for member communications, talk to your compliance team about what can and can't go into an external AI prompt—and consider whether your credit union has a data-handling policy that covers this.

For AI-generated content that reaches your board, your members, or your examiners, a human review step is non-negotiable. Hallucinations are real. An AI can produce a beautifully formatted exam response letter that contains a factual error. The prompting technique improves consistency; it doesn't replace judgment. If your credit union is just getting started—using AI for internal drafts, agenda prep, or marketing copy—the risk is low and the upside is real. Go. If you're moving toward automating member-facing or regulated communications, loop in compliance before you ship anything.

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

Frequently Asked Questions

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

The post Zero-Shot, One-Shot, Few-Shot: How Many Examples Does Your AI Actually Need? appeared first on CU 2.0.

Year in, year out maybe 150 credit unions go out of business, mainly as a result  of mergers, but a few voluntarily close and NCUA sometimes shuts a few doors.

Year in, year out new credit unions spring to life.  Two or three every year.  

Do the math,.  It’s grim.  It’s not a pleasant look into credit union futures.

Now listen to Ben Hering, formerly director of engagement at the Minnesota Credit Union Network, where he helped birth Arise Credit Union. He now is director of CU System Partnerships at consultancy CuCollaborate where his focus is on helping to stimulate the development of de novo credit unions and his near-term goal is to double today’s birth rate.

How? In today’s show he tells how and he also acknowledges the hurdles.

This isn’t an easy task to take on but, as Hering says, he’s helped do it before and he expresses optimism that he can do it again.

He will need help from many corners. Including, notably, vendors to credit unions but also the regulators and politicians.

Is Hering’s a Sisyphean pursuit?  

A cautionary tale is the story of Maine Harvest, a de novo credit union featured in episode 8 of this show.  Hold the applause.  Maine Harvest did successfully launch.  But within a couple years it unfurled a white flag and merged into a bigger Maine credit union.

But, for now, Hering’s is an optimistic voice.  

And many more new credit unions would indeed be a wonderful thing.

The post CU 2.0 Podcast Episode 425 CuCollaborate’s Ben Hering on Birthing New Credit Unions appeared first on CU 2.0.

ASHLAND, Ore. — CU 2.0 today announced expanded roles for two longtime team members as its AI practice grows. Dylan Moore is taking on the role of Head of Product, and both Moore and Head of Marketing Zeke Hudson are serving as coaches in the company’s AI coaching program. Both Moore and Hudson have been with CU 2.0 for more than eight years. That time in the credit union and fintech world shapes how they approach AI, keeping the focus on what works for real institutions and real teams.

“Zeke and Dylan have been doing this work for years, with both fintechs and credit unions alike. They help people see what is possible, from basic skills, to automations, to building custom tools. Expanding their roles into coaching just makes official what they have already been doing, and it shows where we are headed.”

— Kirk Drake, CEO, CU 2.0

Moore has spent several years building out CU 2.0’s AI infrastructure. Much of his work now centers around the CU 2.0 AI Platform, shaping the tools credit unions use and making sure they hold up in real-world use. He also coaches credit unions in the AI program. Moore’s eight years across CU 2.0’s technology, data, and operations work give him a strong understanding of how to use AI to solve real problems.

Hudson leads marketing for CU 2.0 and coaches credit unions in the AI program, where he focuses on turning AI skeptics into confident practitioners. He also helps shape the product. His work centers on the people side of AI: helping a team build the fluency and confidence to use it well and keep building on it.

Both roles are effective immediately. Credit unions interested in working with the CU 2.0 team can connect with the team to see where AI could help and how coaching could support their people.

Additional information is available at cu-2.com/credit-union-ai-coaching/

ABOUT CU 2.0

CU 2.0 is a consultancy embedded in the credit union and financial technology ecosystem, helping credit unions find and adopt the right technology and helping fintechs navigate the credit union market. Based in Ashland, Oregon, CU 2.0 works with institutions across the country through its Fintech Call Program, strategic planning services, and—most recently—AI Coaching and Platform.

The post CU 2.0 Expands Roles for Dylan Moore and Zeke Hudson as Its AI Practice Grows appeared first on CU 2.0.

You've opened ChatGPT or Claude. You stare at the blank text box. You type something like "write me a memo about our new auto loan rates" and hit Enter. The result is… fine. Generic. You could have written it yourself in about the same time.

What if the problem isn't the AI? What if it's the prompt?

Most credit union executives are working with half a tool. They're using AI like a vending machine—drop in a vague request, hope something useful falls out. The smarter move is to let AI help you write the prompt itself. That's meta-prompting, and it changes what you get out of every session.

What Is Meta-Prompting?

Meta-prompting is the practice of asking an AI model to help you build a better prompt before you ask your real question. Instead of guessing how to phrase your request, you describe what you're trying to accomplish and ask the AI to design the instructions that will get you there.

Think of it like briefing a new employee. You wouldn't just say "write something about our HELOC product." You'd explain the audience, the tone, the goal, what to avoid, and what a good result looks like. Meta-prompting is that briefing conversation—except the AI is both the person you're briefing and the one who helps you figure out what to say.

Meta-prompting turns a vague intention into a precise instruction set. It's the difference between telling a contractor "fix the kitchen" and handing them a detailed scope of work—except AI helps you write the scope.

Why Does This Matter for Credit Unions?

Your team uses AI for a wide range of tasks: drafting board memos, summarizing vendor proposals, writing member-facing FAQs, preparing talking points for an examiner visit, comparing core conversion options. Each of those tasks has a different audience, a different risk profile, and a different definition of "good."

A generic prompt produces a generic result. A well-engineered prompt—one that specifies audience, tone, format, constraints, and goal—produces something your lending team or your marketing coordinator can actually use with minimal cleanup.

The challenge is that most executives don't have time to learn prompt engineering from scratch. Meta-prompting solves that. You don't have to know how to write a perfect prompt. You just have to know what you want.

How Do I Use Meta-Prompting? (Step by Step)

Here's a workflow you can use today. No technical background required.

1. Open your AI tool of choice. ChatGPT, Claude, Gemini—this technique works across all of them. For sensitive internal documents, make sure you're using a version your compliance team has approved for credit union data.

2. Describe your goal in plain language. Don't worry about phrasing it perfectly. Tell the AI what you're trying to produce, who it's for, and why. For example: "I need to write a memo to our board explaining why we're moving to a new digital banking platform. The board is financially sophisticated but not tech-savvy. I want them to feel confident, not anxious."

3. Ask the AI to write the prompt for you. After your description, add something like: "Based on that, write me the best possible prompt I should use to generate this memo." The AI will return a detailed, structured prompt—often with role instructions, format guidance, and specific constraints you wouldn't have thought to include.

4. Review the generated prompt carefully. Does it reflect your actual goal? Are there things your compliance team would flag? Add any credit union-specific context the AI missed—your field of membership, your tone guidelines, regulatory sensitivities for your examiner relationships.

5. Run the refined prompt. Paste it into a new chat window (starting fresh keeps the AI from anchoring on your earlier description) and let it work. Compare this result to what your original vague prompt would have produced. The difference is usually significant.

6. Save prompts that land well. When a meta-prompted result produces something useful, save that prompt in a shared document. Your member services team, your loan ops team, your marketing team—they can all reuse it. Over time, you're building a credit union-specific prompt library without any formal engineering effort.

What Does This Look Like in Practice?

Say your compliance officer needs a plain-English summary of a new NCUA rule update for branch staff. Without meta-prompting, someone on the team types: "Explain the new NCUA rule in simple terms." The result is a wall of text with no structure and too much jargon.

With meta-prompting, the process looks like this: describe the task to the AI (NCUA rule update, branch staff audience, plain English, max one page, must cover what changes and what stays the same), ask for the ideal prompt, review it, then run it. The output comes back formatted, appropriately scoped, and written at a level your tellers can actually understand.

That's not magic. That's just better instructions.

The executives who get the most out of AI aren't the ones who type the most. They're the ones who slow down for sixty seconds, describe what they actually need, and let the AI help them engineer the ask.

A Few Things to Keep in Mind

Meta-prompting is a productivity tool, not a compliance shortcut.

The AI doesn't know your credit union. It will make assumptions about your audience, your brand voice, and your regulatory environment. Always review generated prompts before running them and add credit union-specific context where needed.

Sensitive data stays out of public AI tools. If the task involves member information, internal financials, or anything your examiners would raise an eyebrow at, use an enterprise or private AI environment. Consult your compliance team on what's approved for your setup before you start.

Finally, meta-prompts aren't perfect either. The AI might generate a prompt that sounds thorough but misses something critical for your situation. Treat the generated prompt as a strong first draft, not a finished product.

What's the risk?

For most meta-prompting use cases—drafting member communications, writing board memos, summarizing vendor proposals, building examiner prep materials—the risk is low. You're asking AI to help you write instructions, not to process member data or make lending decisions. As long as you're keeping sensitive information out of public AI tools, the downside is mainly that you get a mediocre prompt and have to revise it. That's manageable.

The risk increases when the underlying task involves member NPI, confidential board materials, or regulatory filings. The meta-prompting technique itself is fine for those tasks, but the tool environment matters enormously. If you're at a credit union that hasn't yet established an approved AI environment for sensitive work, talk to your compliance team before putting anything confidential into a chat window—regardless of how you're prompting it. From an NCUA examiner standpoint, AI use in credit unions is still an evolving area; document your internal policies and be ready to explain your controls. Consult your compliance team before using any AI tool in a regulated workflow.

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

Frequently Asked Questions

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

The post Meta-Prompting: How to Ask AI to Write Your Prompt For You appeared first on CU 2.0.

In the first half of 2026 credit union mergers were down, admits Peter Duffy, a mergers expert now with Strategic Resource Management – but don’t expect that to be a permanent downturn.

Mergers will pick up and there will continue to be big credits merging with big credit unions.

Duffy, who has guested on the show multiple times before, is one of the leading experts on the subject of credit union mergers.  On this episode he tells why there’s been a slowing in 2026, and also why it seems an inevitability that we will see renewed interest among credit unions in doing a merger.

It was just a handful of years ago when a credit union merger was something of a rarity and when it did occur it usually was what I have long called a mercy marriage – some might call it a shotgun wedding – which meant a big credit union swallowed a little one at the request of a regulator.

What’s different now is that many – most? – credit unions have begun to see merger as a strategic gambit that just might be the right path into a more successful future.

The post CU 2.0 Podcast Episode 424 Peter Duffy On Credit Union Merger Trends: Buckle Up! appeared first on CU 2.0.

ASHLAND, Ore. — CU 2.0 launched the CU 2.0 AI Platform, giving credit unions a set of AI tools built for their work and ready to put to use. Credit unions choose the tools that fit, deploy them for their teams and members, and CU 2.0 works alongside each institution to tune them to how it operates.

The platform includes AI tools for everyday credit union work such as member service, lending, and compliance, along with member-facing applications a credit union can deploy and governance and NCUA-readiness support its teams can use. It also offers a fast way to pull NCUA data and peer benchmarks, without the usual manual digging. CU 2.0 builds and refines these tools together with the credit unions that use them.

The platform is also built to be shared. When a credit union develops a tool of its own, it can offer it to others on the platform, so every institution draws from a wider, tested set of solutions. It is the credit union model at work: institutions helping each other rather than each building alone.

“Credit unions do their best work together. We built the platform around that. Every credit union gets tools that are ready to use, and a place to share what they build so the next one doesn’t start from zero.”

— Kirk Drake, CEO, CU 2.0

The platform works for credit unions of any size, whether a team is new to AI or already building with it. CU 2.0’s AI Coaching and AI Mastermind programs run alongside it, so teams know how to use the tools well and get more out of them over time.

Credit unions can see the platform and get started at cu-2.com/ai-platform, or book a free AI Discovery Call with the CU 2.0 team to find the tools that fit and put them to work.

Additional information is available at cu-2.com/credit-union-ai-coaching/

ABOUT CU 2.0

CU 2.0 is a consultancy embedded in the credit union and financial technology ecosystem, helping credit unions find and adopt the right technology and helping fintechs navigate the credit union market. Based in Ashland, Oregon, CU 2.0 works with institutions across the country through its Fintech Call Program, strategic planning services, and—most recently—AI Coaching and Platform.

The post CU 2.0 Launches an AI Platform Built for Credit Unions appeared first on CU 2.0.

You've heard of Custom GPTs. Maybe you've even built one. But now Anthropic is talking about "Claude Skills," and you're wondering: is this the same thing with a different name, or something you actually need to pay attention to?

It's something you need to pay attention to. Both tools let you extend an AI assistant beyond its out-of-the-box defaults. But they work differently, live in different places, and are better suited for different jobs inside your credit union. Here's what you need to know to make a smart call about which one belongs in your workflow—and when to use both.

What Is a Claude Skill?

A Claude Skill is a custom capability you add to Claude, Anthropic's AI assistant, that lets it connect to external tools, data sources, or workflows. Think of it as giving Claude a specific job and the equipment to do it.

A standard Claude session is essentially a conversation. A Skill layers in context, instructions, and integrations. You might build a Skill that pulls from your credit union's policy documents, connects to your ticketing system, or knows how to format output for your board report template.

As of 2025, Skills are primarily configured through Claude's API and through Claude for Work—Anthropic's business tier. Verify current availability and pricing with Anthropic directly, as these offerings are evolving quickly.

A Claude Skill isn't just a smarter prompt. It's Claude with a defined role, a rulebook, and access to the right information—purpose-built for a specific job at your credit union.

What Is a Custom GPT?

A Custom GPT is OpenAI's version of the same basic idea: a configured version of ChatGPT that you shape with instructions, a persona, and optionally uploaded documents or connected tools. Custom GPTs live inside ChatGPT and are built through a no-code interface that most executives can figure out in an afternoon.

If you've already built one—say, a GPT that drafts member-facing email campaigns in your credit union's voice, or one that summarizes vendor proposals using your standard evaluation framework—you know how quickly they become genuine time-savers. For a deeper look at building them, see our post on building Custom GPTs for credit unions.

How Are Claude Skills and Custom GPTs Different?

They share the same concept but differ on three dimensions that matter in a credit union environment: flexibility, technical lift, and model behavior.

Flexibility. Claude Skills, particularly when built through the API, offer deeper integration with external systems. A Skill can reach into a connected data source—your loan pipeline tool, your member feedback platform, your internal knowledge base—and pull live information into the response. Custom GPTs can also connect to external tools via Actions, but many credit union teams find the Claude integration model more straightforward for complex back-end connections.

Technical lift. Custom GPTs win on accessibility. Any executive with a ChatGPT Plus or Team account (verify current pricing) can build one through a guided interface—no code required. Claude Skills at their most powerful require API access and someone comfortable with technical configurations. That said, Claude for Work is lowering that bar, and the gap is narrowing.

Model behavior. Claude and GPT-4 simply behave differently. Many credit union teams that handle sensitive member data or draft compliance-adjacent content find Claude's responses feel more careful and measured. That's not a universal truth—it's a preference worth testing with your own team on your own use cases.

How Do I Build a Claude Skill for My Credit Union?

If you have access to Claude for Work or your team has API capabilities, here's a practical path forward.

1. Pick one narrow problem. Don't try to build a Swiss Army knife on day one. Choose a single, repeatable task—something like summarizing examiner findings into a draft response memo, or converting raw member survey data into a formatted board presentation.

2. Write the system prompt. This is the instruction set Claude reads before every session. Be specific: tell it your credit union's name, the role it's playing, the format you want, and any guardrails. For example: "Never speculate about regulatory timelines; flag those for the compliance team."

3. Load your documents. If your Skill needs to reference internal policies, product guides, or past board reports, upload them or connect them through your integration layer. This is where Skills start to outperform a generic AI session.

4. Test with real scenarios. Run three to five actual tasks through it. A good test for a loan ops Skill: paste in a recent underwriting summary and see if the output matches what your team would actually send up the chain.

5. Set access controls. Decide who on your team can use it and confirm that your data-handling setup aligns with your information security policy. Loop in your compliance team before rolling it out broadly—especially if the Skill touches member data.

6. Iterate. The first version will be imperfect. Build in a two-week feedback loop with the team members using it daily.

When Should You Use a Skill Instead of a Custom GPT?

Use a Custom GPT when you need something fast, accessible to non-technical staff, and self-contained. Great candidates: member communication templates, meeting agenda builders, job description drafts, or a board packet summarizer that anyone on your exec team can run without IT support.

Reach for a Claude Skill when the job requires live data connections, more complex workflows, or when your team already works inside Anthropic's tooling. Also consider it if head-to-head testing shows Claude's output quality is a better fit for a particular task—drafting sensitive member communications, for instance, or producing first drafts of policy documentation.

Many credit unions will end up using both. That's not a failure to pick a winner. That's using the right tool for the right job.

Why Does This Matter for Credit Unions Specifically?

Credit unions operate under a level of regulatory scrutiny and member trust that most businesses don't. When you configure an AI tool—whether it's a Skill or a Custom GPT—you're making decisions about what information it can access, what it can say on your behalf, and who can use it.

Both platforms are improving their enterprise controls quickly. But the stakes of getting it wrong are real: a compliance gap surfaced by an examiner, a member-facing output that creates fair lending exposure, a data-handling misstep. You want to be deliberate here.

The good news is you don't have to figure this out from scratch. Both OpenAI and Anthropic publish their enterprise data handling terms. Your core and digital banking vendors are increasingly building AI integrations directly. And the credit union community is actively sharing what's working.

Start small. Test rigorously. Bring your compliance team in early—not after you've already deployed.

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

What's the risk?

The risk profile here depends entirely on what you connect these tools to. A Claude Skill or Custom GPT that helps your marketing team draft social copy or your HR team write job descriptions? Low risk. Go. But the moment either tool touches member NPI—loan data, account information, contact records—you're in a different conversation. Both OpenAI and Anthropic publish enterprise data handling terms, and those terms matter to your examiners. Review them with your information security officer and compliance team before you build anything that integrates with your core, your loan origination system, or your digital banking platform. Consult your compliance team for guidance specific to your charter and regulatory environment.

There's also a subtler risk: over-reliance on AI-generated output in compliance-adjacent work. A Skill that drafts examiner response memos is genuinely useful—but someone with real regulatory knowledge still needs to review every word before it goes out. AI hallucinations are less common than they used to be, but they haven't disappeared. Build human review into any workflow that carries regulatory or member-facing consequences. The tool is the assistant; your team is still accountable.

Frequently Asked Questions

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

The post Claude Skills vs. Custom GPTs: What Credit Union Executives Need to Know appeared first on CU 2.0.

We need more business members!

That could count as a credit union chant of the last decade but a reality is that many credit unions have simply not achieved their goals in terms of enrolling more business members.

Enter Goodbuy, which describes itself as “the game changing solution for credit unions, accelerating new SMB account openings.”

Now know this: Boise based Goodbuy in June registered as a CUSO and in the process it closed on a $1 million investment from One Washington Financial (OWF), Reseda Group (a wholly owned CUSO of Michigan State University Federal Credit Union), Maps Credit Union and Together Credit Union.

One Washington Financial itself is a CUSO, wholly owned by Washington State Employees Credit Union and its mission is investing in the future of credit union innovation.

On this show is Scott Daukas, chief partnership officer at One Washington Financial and himself a past guest on this podcast.

Daukas explains why One Washington Financial invested in Goodbuy – WSECU itself is implementing the Goodbuy platform – and he adds that he envisions very good things for credit unions that work with Goodbuy.

Also on the show is Goodbuy CEO Cara Oppenheimer who says that Goodbuy is a credit union-branded community marketplace that helps credit unions offer exclusive local business discounts to members, creating a unique differentiator against traditional banks while driving deposit growth and engagement.

Sound good?  

Oppenheimer tells the details – and she details Goodbuy successes with Michigan State University Federal Credit Union which has launched its Hello, Local which helps local businesses offer special deals to MSUFCU members, making this a win-win-win for the business, the consumer member and the credit union.

The post CU 2.0 Podcast Episode 423 Meet Goodbuy the Innovative CUSO That Links Small Businesses to CU Members appeared first on CU 2.0.

Your VP of Lending swears by ChatGPT. Your marketing coordinator uses Gemini. Someone on the IT team just sent a Slack message about Perplexity. And your Microsoft rep is calling—again—about Copilot.

So which one should your credit union actually use?

The short answer: it depends on what you're trying to do. But that's not a dodge. These tools have genuinely different strengths, and picking the wrong one for the wrong job costs you time, money, and sometimes trust. Here's a plain-English breakdown so you can make a smarter call.

What Are These Tools, Really?

All five are AI assistants built on large language models. You type something in; they respond in natural language. But they're built on different underlying models, trained on different data, woven into different software ecosystems, and priced differently.

Think of them less like interchangeable apps and more like five contractors who all know how to build—but each has a specialty, a day rate, and a preferred tool belt.

ChatGPT: The Swiss Army Knife

Best for: General drafting, brainstorming, member-facing content, custom workflows

ChatGPT, made by OpenAI, is the name most people think of when they hear "AI." It's versatile, widely documented, and has the largest community of users sharing prompts and use cases.

For credit unions, ChatGPT shines when your team needs to produce content quickly: member newsletters, board memo summaries, job postings, social media copy. The Custom GPTs feature (available on paid plans, as of 2025) lets you build a version of ChatGPT trained on your own instructions, tone guidelines, and uploaded documents. Your member services team could use a Custom GPT pre-loaded with your product FAQs so every response stays on-brand and accurate.

The main limitation: ChatGPT doesn't connect to your live systems. It works on what you give it in the conversation window or pre-load into a Custom GPT. It is not pulling data from your core in real time.

Pricing: Free tier available. ChatGPT Plus runs approximately $20 per user per month. Team and Enterprise plans are available—verify current pricing at openai.com.

Claude: The Careful Thinker

Best for: Long documents, policy analysis, board materials, anything requiring nuance

Claude, made by Anthropic, handles long documents exceptionally well. You can paste in an entire vendor contract or a 40-page NCUA exam report and ask it to summarize key risks, flag unusual language, or draft a response memo. Many credit union executives who use both ChatGPT and Claude describe Claude as "the one you trust with the serious stuff."

Claude tends to reason through complexity rather than pattern-match to a quick answer. For credit union work involving regulatory language, vendor agreements, or board presentations, that extra care matters.

Claude is also generally seen as more cautious about making things up—though no AI tool is immune to errors. Always verify specific claims before acting on them.

Claude offers an enterprise tier with stronger data privacy commitments. If your board or examiners ask how you're protecting member data when using AI, Anthropic's data-use terms are worth reviewing with your compliance team before you expand use.

Pricing: Free tier available. Claude Pro runs approximately $20 per user per month (as of 2025)—verify current pricing at anthropic.com.

Gemini: The Google Integration Play

Best for: Teams already living in Google Workspace—Docs, Sheets, Gmail, Meet

Gemini is Google's AI, and its strongest argument is ecosystem fit. If your credit union runs on Google Workspace, Gemini integrates directly into the tools your team already uses. Draft a loan officer job description in Google Docs with Gemini's help. Summarize a long email thread in Gmail. Build a quick data model in Sheets.

For your marketing team or ops team that spends the workday inside Google's suite, Gemini can reduce friction because you're not switching apps. The AI is already inside the document you're working on.

The limitation is similar to ChatGPT's: it doesn't have live access to your core, your digital banking platform, or your member data. It works within the scope of what's in your Google environment.

Pricing: Gemini is included in some Google Workspace tiers; Gemini Advanced is available as an add-on. Verify current pricing at workspace.google.com.

Microsoft Copilot: The Microsoft 365 Bet

Best for: Teams already on Microsoft 365—Word, Excel, Teams, Outlook, SharePoint

Copilot is Microsoft's answer to Gemini, built on OpenAI's models and woven into Microsoft 365. If your credit union runs on Teams, Outlook, SharePoint, and Word, Copilot can be genuinely useful: summarize a Teams meeting you missed, draft a Word document from a bullet-pointed outline, pull insights from a SharePoint-stored policy document.

One real advantage for larger credit unions: Copilot for Microsoft 365 can reference documents stored across your Microsoft environment. In practice, that means it can draft a memo and pull context from your SharePoint policy library without you copying and pasting.

The cost has historically been higher than the other tools on this list. Early enterprise pricing ran approximately $30 per user per month on top of existing Microsoft 365 licensing (as of 2025)—verify current pricing at microsoft.com. For smaller credit unions with tighter budgets, that math may not pencil out right now.

Perplexity: The Research Tool

Best for: Current-events research, market scans, quick factual lookups with citations

Perplexity is different from the others. It's less of a "help me write" tool and more of a "help me know" tool. It's an AI-powered search engine that pulls from live web sources and shows you citations alongside its answers.

For credit union executives, Perplexity is useful for three kinds of tasks:

1. Scanning competitor rates. Ask what the average high-yield savings rate looks like today and it pulls current data from live sources.
2. Researching a vendor. Get a quick overview of a fintech you're about to meet with—recent news, funding rounds, customer reviews.
3. Following regulatory updates. Search for recent NCUA guidance or CFPB rulemaking commentary without wading through search results manually.

It's not the right tool for drafting a board memo or building a member-facing chatbot. But as a research accelerator for your leadership team, it earns a seat at the table.

Pricing: Free tier available. Perplexity Pro runs approximately $20 per month (as of 2025)—verify current pricing at perplexity.ai.

How Do You Actually Pick One?

Start with the questions your team is trying to answer most often. Then work through this sequence:

1. Audit your most common AI use cases. Ask three or four department heads: "What tasks do you wish took less time?" Loan officer letter drafting, exam prep summaries, member complaint responses, board report formatting—these are all legitimate AI candidates.
2. Check your existing software contracts. If you're already paying for Microsoft 365 or Google Workspace at scale, the built-in AI tools may cost far less than adding a standalone subscription. Talk to your current reps before buying something new.
3. Run a 30-day pilot with one team. Don't roll out five tools at once. Pick one use case, one department, one tool. Measure time saved and quality of output. Then decide whether to expand.
4. Get your compliance team involved early. Any AI tool that touches member data—even indirectly—warrants a review of the vendor's data handling policies. Consult your compliance team before expanding use beyond internal, non-member-data tasks.
5. Set a simple acceptable-use policy. Before your staff experiments on their own, give them a one-page guide on what's allowed. Member data does not go into free-tier AI tools. Anything AI-generated gets a human review before it goes out.

A Quick Side-by-Side

No single tool wins across every category. Many credit unions end up using two: one for drafting and one for research. That's a reasonable place to land.

What's the risk?

The risk profile here varies by tool and by how you use it. The biggest exposure isn't the AI itself—it's what your staff puts into it. Member names, account numbers, loan details, and any other non-public personal information should never go into a free consumer-tier tool. Most major providers offer enterprise or team plans with stronger data handling commitments and, often, a promise not to train their models on your inputs. That matters a great deal if examiners ever ask how you're managing AI-related data risk. Before you expand any of these tools beyond internal, non-member-data tasks, loop in your compliance team and review the vendor's data processing terms. Consult your legal counsel on anything that touches member NPI.

The second risk is accuracy. Every tool on this list will occasionally produce confident-sounding information that is simply wrong. For low-stakes work—marketing copy, internal training material, first drafts of meeting agendas—the risk is low and the upside is real. For anything that informs a regulatory response, a board decision, or a member-facing policy, treat AI output as a starting point, not a finished product. A human reviewer who knows your credit union's situation needs to be in the loop before anything goes out the door. On vendor concentration: if your team becomes deeply dependent on a single provider, watch for pricing changes and have a rough contingency plan. The AI market is moving fast, and today's pricing is not guaranteed tomorrow.

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

Frequently Asked Questions

This post was drafted with AI assistance and reviewed by a human at CU 2.0. AI makes mistakes; verify any specific claim before acting on it.

The post ChatGPT vs. Claude vs. Gemini vs. Copilot vs. Perplexity: Which AI Tool Should Your Credit Union Actually Use? appeared first on CU 2.0.