10 Biggest Credit Union AI Wins of 2025 | 12/24

Plus: Commonwealth CU automates 70%+ of consumer loan decisions, WEOKIE FCU automates 66% of calls, and MUCH more!

We’ve been watching AI adoption across the CU space all 2025.

A LOT has happened this year… some really big news, some less.

So, I picked the biggest AI wins with real, tangible results:

1. Commonwealth CU automates 70%+ of consumer loan decisions

2. WEOKIE FCU saves $800K per year by automating 66% of calls

3. Heritage FCU drives 2.5x loan growth

4. Capitol CU of Texas grows new membership 59%

5. Nutmeg State FCU automates 97.6% of back-office work

6. Sound CU saves $15K+ per month with 24/7 AI self-service

7. Towpath CU grows revenue 15% YoY

8. 3Rivers FCU takes down 250+ phishing sites in minutes

9. Verity CU lifts approvals 235% across protected classes

10. First City CU cuts early-stage collections costs 54%

Read time: 9 minutes

Top Stories of 2025

The biggest news this year…

1) Commonwealth Credit Union automates 70%+ of consumer loan decisions while cutting delinquency 30%

Commonwealth Credit Union, a $2.2B CU serving 120,000 members, needed to lend faster and more consistently while managing risk. Meanwhile, experienced underwriters became harder to retain, and member demand for instant decisions increased.

After modernizing its lending decisioning, Commonwealth saw:

• 70-83% of consumer loan decisions automated

• 30-40% lower delinquency compared to traditional scoring models

• 340% increase in automation for auto and personal loan decisions

• $324M+ in consumer loans approved across 18,000+ loans since 2021

What changed behind the scenes: Commonwealth implemented Zest AI underwriting across auto, credit card, and personal loans, integrated AI decisioning into its digital application flows, and expanded automation to pre-screen, pre-approve, and manage risk at scale. (link)

2) WEOKIE Federal Credit Union saves $800K per year by automating 66% of calls

WEOKIE Federal Credit Union, with 64,000 members, was facing surging call volumes, up to 30-minute hold times, and rising agent burnout just as its legacy phone banking system was being discontinued. With 20,000-25,000 calls per month and high staff attrition, expanding headcount wasn’t sustainable. But member experience was at risk.

After upgrading from IVR to AI-powered voice automation, WEOKIE achieved:

• 66%+ of incoming calls automated

• Member wait times reduced to under 30 seconds

• 9,000+ after-hours calls handled per month

• $800,000 saved annually in staffing and operational costs

• Significant reduction in agent burnout and turnover

• 24/7 member support without adding headcount

To drive those results, WEOKIE rolled out a phased Voice AI deployment with interface.ai that automated high-frequency inquiries, integrated securely with core systems, and routed complex issues to agents with full context. (link)

3) Heritage Federal Credit Union drives 2.5x loan growth after exceeding 151% of its annual target

Heritage Federal Credit Union, with ~$1B AUM and 65,000 members, was constrained by legacy contact center tech that left agents buried in low-value calls, limited service-level improvements, and made scaling growth without hiring nearly impossible. The contact center was operating defensively, focused on keeping up rather than driving revenue.

After shifting to an AI-powered contact center model, Heritage unlocked:

• 2.5x year-over-year increase in loan volume

• 151%+ of annual loan growth target achieved

• 2x more member volume handled with no additional headcount

• 20% improvement in workforce efficiency

• 100% of member interactions automated or AI-enhanced

• 18% reduction in average handle time

• 16% increase in service levels

So what did they actually do? Heritage replaced its legacy CCaaS with Glia Voice and AI agent Heidi. Then, they deployed Glia Digital across web, SMS, digital banking, and voice, using AI to deflect calls and free up agent time. (link)

4) Capitol Credit Union of Texas grows new membership 59% and makes digital its #1 account-opening channel in 90 days

Capitol Credit Union of Texas needed to move beyond an aging membership and compete for younger, digital-first members, but its existing account-opening experience couldn’t scale or show which marketing investments were actually working.

Once digital account origination was rebuilt around automation and analytics, Capitol saw:

• 59% increase in new memberships (vs. prior three months)

• ~$10,000 in average new deposits per new member

• 70% reduction in employee processing time

• Digital became the #1 channel for new accounts within 90 days

• Faster onboarding with improved fraud controls

Behind the lift, Capitol got rid of its legacy account-opening stack for Cotribute’s digital account origination and decision intelligence platform. Now, back-office reviews are automated, Jack Henry Symitar is integrated, and analytics are used to scale growth. (link)

5) Nutmeg State Financial Credit Union automates 97.6% of back-office work and enables instant decisions for 89.5% of applications

Nutmeg State Financial Credit Union was dealing with rising fraud attempts, low conversion rates, and heavy operational strain from manual reviews that slowed decisions and overburdened staff. As application volume grew, the existing process increased both cost and error risk.

Once account and loan processing was restructured around rules-based decisioning, Nutmeg State saw:

• 97.6% of manual back-office processing steps automated

• 89.5% of account opening applications decisioned instantly

• Fewer fraudulent applications reaching staff review queues

Behind the scenes, Nutmeg State moved high-volume loan, membership, and deposit applications onto Cotribute’s decision intelligence platform. Cotribute uses 40+ rules to route applications into auto-approval, auto-denial, or manual review queues, reducing reliance on manual processing. (link)

6) Sound Credit Union saves $15K+ per month by launching 24/7 AI self-service

Sound Credit Union, with $3B in assets and 26 full-service branches, was struggling with inconsistent answers, rising chat abandonment, and limited staffing that made it risky to expand into mobile chat without hurting service levels. As digital demand grew, the Credit Union needed a way to scale support across channels without adding headcount.

After rolling out AI-powered self-service, Sound CU achieved measurable gains:

• 77% of member inquiries fully resolved without a live agent

• Wait times cut from 1.5 minutes to 19 seconds

• Chat abandonment reduced from 4.8% to 1.3%

• 3.5 FTE per month avoided, saving $15K+ in monthly staff costs

• Member satisfaction jumped from 4.0 to 4.4 out of 5

Sound CU achieved these results by deploying Posh across web, online banking, and mobile banking using prebuilt AI topics and a single unified platform that requires under two hours per month to maintain. (link)

7) Towpath Credit Union grows revenue 15% year over year after a one-day digital banking overhaul

Towpath Credit Union, a $205M Credit Union serving 25,000 members, was held back by digital gaps that risked frustrating members, slowing growth, and limiting engagement. Leadership needed a modern digital foundation that could scale relationships beyond just transactions.

Following a full-stack digital conversion, Towpath saw:

• 15% year-over-year revenue growth

• 8.7% year-over-year loan growth

• 9.5% year-to-date growth in new account holders

• Digital banking adoption increase from 40% to 57%

• Products per member growth from 1.7 to 2.4

• Efficiency ratio improvement from 83% to 81%

Towpath completed a one-day, full digital conversion with Alkami, centralizing digital banking, onboarding, data-driven marketing, and financial wellness tools into a single platform. Then, they used member data to drive engagement and recapture wallet share. (link)

8) 3Rivers Federal Credit Union takes down 250+ phishing sites in minutes with automated fraud response

3Rivers Federal Credit Union, with $2.47B in assets and 106,000 members, was dealing with a surge in phishing attacks that led to account takeovers and heavy manual work to remove fake websites and mobile apps. The fraud team needed a faster way to neutralize threats without tying up staff.

With automated phishing detection and takedowns in place, 3Rivers reported:

• 250+ phishing sites, fake apps, and social accounts taken down

• Takedowns typically completed in under 15 minutes

• Reduced manual workload for fraud and digital teams

• Fewer active phishing threats targeting members

• Improved member communication around fraud trends

3Rivers centralized phishing reporting through Alkami and routed takedowns to Appgate, allowing suspicious sites to be flagged, ticketed, and removed quickly while using reporting data to proactively educate members. (link)

9) Verity Credit Union lifts approvals 235% across protected classes with inclusive AI underwriting

Verity Credit Union, an $859M CDFI with 40,000 members, was constrained by generic credit scoring models that excluded qualified borrowers, introduced bias into decisioning, and limited its ability to expand inclusive lending programs. To stay true to its “People, Planet, Prosperity” mission, Verity needed faster, fairer underwriting without increasing risk.

With AI underwriting in place, Verity reported:

• 235% average jump in approvals across protected classes

• 84% more instant credit card approvals

• 100% more instant auto loan approvals

• 69% more in instant personal loan approvals

• Significant approval gains across underserved segments, including seniors (271%), women (194%), and minority borrowers (158%-375%)

The key: Verity replaced generic credit scoring with Zest AI underwriting across auto, personal loan, and credit card portfolios, embedding AI decisioning directly into its LOS. (link)

10) First City Credit Union cuts early-stage collections costs 54% by shifting reminders from calls to automated texts

First City Credit Union, with $900M in assets and 56,000 members, was spending nearly $90,000 per year on third-party collection calls for early-stage delinquencies. The calls were costly, uncomfortable for members, and pulled staff away from higher-value recovery work.

Once early-stage collections moved from phone calls to automated texting, First City achieved:

• 54% reduction in third-party collections costs

• ~$48,000 in annual savings

• Very low opt-out rates from members

• Increase in member compliments to loan servicing staff

• Collectors freed up to focus on complex recoveries

First City shifted early-stage collections to automated text campaigns using Eltropy and integrated those workflows with Akuvo, triggering reminders based on delinquency status and embedding payment links directly into messages. (link)

Tips & Use Cases

Learn to apply AI…

Reseda Group partners with Constant AI to automate loan servicing: The CUSO is expanding its fintech marketplace for Credit Unions with digital self-service for loan requests like payoff quotes, deferrals, and due date changes. Constant AI says 30-40% of loan operations capacity can be automated. (link)

4 Credit Union AI priorities for 2026: Credit Unions must move from isolated AI pilots to enterprise strategy, focusing on algorithmic lending, back-office automation, Gen Z onboarding, and AI-driven fraud defense. Efficiency and member trust must scale together in an AI-first operating model. (link)

6 traits of the 2026 Credit Union member: Members will be more anxious, more selective, and less tolerant of friction, expecting personalization, simplicity, and real human guidance alongside digital convenience. To stay relevant, Credit Unions must simplify experiences, act hyperlocal, and build true relationships. (link)

Why agentic AI adoption stalls without trust: Even as interest grows, 98% of product leaders aren’t ready to give AI agents access to core systems, and 75% cite governance and security risks. Reliable, well-controlled automation (not budget) is the real prerequisite for trusted AI decision-making. (link)

Firms that use AI for client-facing value will beat those that only focus on efficiency: You can use AI to cut costs and run leaner, or you can productize it to deliver client-facing insights, personalization, and new revenue. Firms that only optimize operations may survive, but those that turn AI into something clients value are the ones positioned to grow. (link)

How to build a budgeting review app with ChatGPT 5.2 in 5 minutes: Turn budget reviews from slow, messy meetings into fast decisions by using this ChatGPT prompt to structure assumptions, surface tradeoffs, and flag risks instantly. (link)

Barclays tests AI platform to turn sustainability into profit insights: Starting in early 2026, Barclays will pilot ExpectAI’s Una platform, which builds digital twins that model how a small or medium-sized business operates so leaders can test energy and cost changes before acting. (link)

Bunq upgrades its AI assistant to handle member support: The neobank’s upgraded GenAI assistant, Finn, now handles about 97% of support interactions using real-time speech-to-speech, answering everyday questions like monthly spending instantly. (link)

BBVA launches “The Eight” to embed AI across the entire bank: The roadmap includes six AI “robots” covering customer service, relationship managers, risk, operations, software development, and employee productivity, supported by shared data and AI architecture. The strategy moves beyond isolated automation toward AI that anticipates needs, personalizes at scale, and improves decisions across the organization. (link)

How CX and compliance can converge to fight attrition: About 13% of retail bank customers are considering switching primary institutions, often through “soft switching” that goes unnoticed. Use AI to turn compliance functions like fraud prevention, identity verification, and complaint analysis into proactive CX advantages that build trust and reduce churn. (link)

Dext finds businesses are losing money by using general AI for financial advice: A survey of 500 UK accountants found that 50% have seen clients lose money due to bad AI advice. 77% have seen increased use of AI tools, and 72% have seen clients challenge professional guidance with AI. Meanwhile, 40% spend 4-10 hours a month fixing AI errors. (link)

Cornerstone Advisors says 2026 will push AI into vendor workflows: 2025 focused on conversions and core modernization, but 2026 will be defined by AI embedded directly inside core, payments, lending, and workflow vendors. The real AI strategy will come less from in-house systems and more from how effectively vendors bake AI into everyday systems. (link)

Credit Union fintech leaders say AI, real-time payments, and embedded lending defined 2025 and set 2026 priorities: In 2025, Credit Unions moved AI from experiments into lending, underwriting, and back-office workflows while pushing real-time payments and embedded finance. Heading into 2026, the focus shifts to AI-first strategies, hyper-personalization, trusted fintech partnerships, and balancing automation with human oversight. (link)

Manual AI compliance is becoming a growth bottleneck: As AI use accelerates, 95% of financial institutions report rising model risk requirements, yet 60% still rely on manual compliance. Model documentation involves an average of 29 people per institution, sometimes more than 50, while 87% plan to adopt automated model documentation within two years to keep pace with regulation and innovation. (link)

Funding Spotlight

Where the money is flowing for innovation…

Thread Bancorp raises $30.5M to scale embedded banking model: The parent of Thread Bank closed a $27.6M round plus a $2.9M extension led by Portage Ventures to expand its embedded banking and digital branch strategy. Thread now manages over $2B in partner deposits and saw roughly 75% annualized growth in embedded deposits. (link)

OnCorps AI raises $55M growth round for agentic fund operations automation: The platform autonomously investigates and resolves exceptions across trade reconciliations, fund accounting, and financial reporting reviews using AI agents trained on millions of real-world discrepancies. Firms representing $13T in assets use the system to cut manual effort as operating margins compress. (link)

Adaptive Security raises $81M Series B to counter AI-driven impersonation attacks: The funding led by Bain Capital Ventures backs expansion of a human-layer security platform that detects and mitigates deepfake voice, video, text scams, and AI-powered phishing that bypass traditional controls. Adaptive has grown to 500+ enterprise customers in under a year and reports a 94 NPS. (link)

FINNY AI closes Series A to automate financial advisor prospecting with AI: The platform uses AI to identify high-intent prospects, automate outreach, and prioritize leads using money-in-motion signals like liquidity events and career changes. FINNY says advisors generate an average of $7.7M in new client assets annually with near-zero acquisition costs. (link)

Keeping up with Tech

The latest in fintech and tools…

The state of consumer AI usage: Despite dozens of launches in 2025, usage and spending are concentrating around one primary AI per user. ChatGPT leads with roughly 800-900M weekly users, while Gemini is growing fast but still trails on engagement and retention. (link)

BlackLine acquires WiseLayer to embed AI agents into finance operations: BlackLine is acquiring WiseLayer to integrate its AI accounting agents, which automate judgment-heavy tasks such as accruals, revenue recognition, and payroll. (link)

Ani Tech launches AI command center to run financial advisor workflows: Ani Pulse, an AI-driven command center that sits on top of existing CRMs to proactively prioritize daily actions, automate admin tasks, and monitor firm health. The system replaces static dashboards with autonomous, real-time workflow orchestration. (link)

OnePay joins Google’s AP2 to secure agentic AI payments: OnePay joined Google’s Agent Payments Protocol to enable AI agents to transact with clear user authorization, secure credential reuse, and transparent financing options. The move aims to keep AI-led commerce aligned with human intent and control. (link)

Visa and Aldar complete first live agentic payment using AI: Visa and Aldar executed a real-world AI-led payment, allowing customers to pay service charges via a voice-enabled AI agent using tokenized Visa credentials. (link)

TCS upgrades BaNCS core with built-in AI agents: Tata Consultancy Services launched BaNCS AI Compass, embedding machine learning, generative AI, and pre-built agents directly into its banking core. The upgrade lets institutions deploy explainable, governed AI across onboarding, underwriting, servicing, and risk workflows without rebuilding systems. (link)

PayPal positions itself as the infrastructure layer for AI shopping: PayPal is building a single integration that lets merchants sell through AI assistants like ChatGPT and Gemini without managing multiple protocols. (link)

Google rolls out Gemini 3 Flash to bring faster AI at lower cost: Gemini 3 Flash is a new model that combines frontier-level reasoning with significantly faster speed and lower pricing. It’s now the default model across the Gemini app, Search AI Mode, and developer platforms. (link)

OpenAI releases GPT-5.2-Codex for long-horizon, agentic coding: The new Codex model improves large refactors, migrations, and real-world software engineering with stronger context handling and cybersecurity performance. (link)

OpenAI strengthens security for ChatGPT’s web-browsing AI: ChatGPT Atlas, OpenAI’s feature that lets an AI agent browse websites and take actions online on a user’s behalf, updated to better block hidden instructions on webpages or emails that could trick the AI into doing the wrong thing. (link)

In Other News

Related news you can learn from…

• ChatGPT mobile app consumer spending tops $3B (link)

• AI could displace banking jobs without causing mass unemployment (link)

• Anthropic publishes public AI risk framework to comply with California (link)

• NIST releases AI cybersecurity profile to align AI risk with existing security frameworks (link)

• AI transformation in financial services will hinge on agents, talent, and governance in 2026 (link)

Community Corner

Memes and visuals…

Thanks for reading!

Until next week,

— Credit Union AI Guy

Enjoying this newsletter? Forward it to a colleague who might appreciate practical AI insights.

Have a question or want to share your own AI experiment? Reply to this email or connect with us on LinkedIn. We’re building a community of curious Credit Union professionals. Come join the conversation!