If agentic AI is the engine of banking's next chapter, lending is the proving ground. Credit decisioning is where artificial intelligence meets the oldest function in banking, the evaluation of risk and the allocation of capital and the collision is producing results that should command the attention of every bank leader in the country. Accenture's 2026 Banking Technology Trends report, documents that AI-first credit systems can increase automated approvals by roughly 50 percent and overall decisioning throughput by 70 to 90 percent. Those are not incremental improvements. They represent a fundamental restructuring of how banks originate, underwrite, and manage credit risk. In my two decades of consulting at CCG Catalyst, I have watched lending technology evolve from paper applications to online forms to mobile origination. What is happening now is different in kind, not just degree. The institutions that harness it will redefine what it means to be a lender. Those that don't will find themselves competing for the loans that algorithms at other institutions have already declined.

AI Underwriting Revolution Is Already Here

The data from the research is unmistakable. EY's AI Bank white paper documents case studies showing 80 percent reductions in manual data entry during loan origination, 50 percent reductions in approval cycle times, and 15 percent reductions in overall review time. Experian reports that 70 percent of organizations will be using composite AI, a blend of generative, predictive, and agentic models by the end of 2026, with lending and credit risk among the primary applications. Finastra's State of the Nation survey found that 49 percent of banks are already using AI to accelerate lending processing. The question is no longer whether AI will transform lending. The question is whether your institution is transforming fast enough.

My concern is not with the institutions that are moving aggressively on AI-powered lending. It is with the ones that are moving cautiously for the wrong reasons. There is a meaningful difference between proceeding carefully because you are building proper governance frameworks, model validation, bias testing, explainability documentation and proceeding slowly because you are uncertain, underfunded, or organizationally unable to change. The first posture is responsible banking. The second is a competitive death sentence disguised as prudence.

The regulatory framework for AI in lending is more established than many bankers realize. The model risk management expectations in SR 11-7, reinforced by OCC Bulletin 2025-26 for community banks, provide a clear governance structure. The CFPB's guidance on adverse action notices in AI-driven credit decisions is explicit: lenders must provide specific, accurate reasons when denying credit, and generic explanations do not satisfy the requirement simply because an algorithm made the decision. The CFPB's Winter 2025 Supervisory Highlights put it plainly "there is no advanced technology exception to federal consumer financial laws". I agree with that principle entirely. Banks that build their AI lending capabilities within this regulatory framework, with rigorous model validation, disparate impact testing, and transparent adverse action processes will move faster in the long run than those trying to figure out governance after deployment.

Private Credit Challenge Banks Cannot Ignore

While banks deliberate on AI underwriting, the competitive landscape is shifting beneath them. Capgemini's World CIB Report documents that global private credit assets have reached $1.7 trillion, with private debt funding growing from 68 percent of global buyout deals in 2021 to 77 percent in 2024. Seventy percent of non-bank lenders now cite private credit as their top revenue driver. This is not a niche market segment. It is a fundamental reallocation of credit intermediation away from the banking system and toward private capital.

My view is that banks have brought much of this on themselves. Years of post-crisis regulatory caution, combined with rigid approval processes and slow decisioning timelines, created a vacuum that private credit was designed to fill. A middle-market company that needs a $50 million credit facility does not want to wait six weeks for a committee decision when a private credit fund can provide a term sheet in ten days. The largest banks have recognized this reality, several have committed $50 billion or more to direct lending platforms and co-origination partnerships but community and regional banks are largely absent from this conversation, and they shouldn't be. The commercial relationships that smaller institutions rely on are vulnerable to private credit competitors who offer speed, flexibility, and structuring creativity that traditional bank processes cannot match.

The answer is not to abandon credit discipline. It is to modernize the credit process so that discipline and speed are no longer in opposition. AI-powered credit analysis, automated financial spreading, real-time portfolio monitoring, and dynamic risk pricing can compress decision timelines from weeks to days without sacrificing the analytical rigor that separates sound lending from reckless lending. The institutions I advise that are implementing these capabilities are not cutting corners on credit quality. They are eliminating the manual bottlenecks that add time without adding insight.

Embedded Lending and the Point-of-Need Revolution

The research from 11:FS on embedded banking is particularly relevant to lending. When credit is embedded at the point of need, a small business applying for inventory financing within their accounting software, a consumer accessing an installment option at checkout, a contractor securing equipment financing through a dealer management system, conversion rates increase by 20 to 30 percent. The embedded lending market is projected to reach $370 billion by 2035 and buy-now-pay-later alone is expected to hit $437 billion globally by 2027.

For banks, the strategic question is whether they will be the capital behind these embedded lending experiences or whether they will cede that role to fintech lenders and private credit funds. I believe community and regional banks have a significant opportunity here, but only if they invest in the API infrastructure, real-time decisioning engines, and partner management capabilities required to participate. The 2023 interagency third-party risk management guidance applies squarely to these partnerships, and the consent orders against BaaS banks in 2024 underscore the regulatory expectation that the bank of record maintains genuine oversight regardless of who originates the customer interaction. Embedded lending done right with the bank maintaining credit discipline, compliance oversight, and portfolio governance is an extension of relationship banking into new channels. Done poorly, it is a compliance disaster waiting to happen.

Small Business Lending Opportunity

BAI's 2026 Banking Outlook identifies small business lending as a $130 billion revenue opportunity across 34.8 million small businesses, with $242 billion in annual unsecured SMB lending. Community banks and credit unions are uniquely positioned to serve this market, they understand local business conditions, they have existing relationships, and they have the trust that comes from decades of community presence. But they are losing ground to fintech lenders who offer faster decisions, simpler applications, and digital-first experiences that meet small business owners where they are.

The typical small business owner does not want to gather three years of tax returns, compile personal financial statements, and sit through multiple meetings for a $75,000 working capital line. They want to connect their accounting software, authorize a bank's data pull, and receive a decision within hours. The technology to deliver this experience exists today. The institutions that deploy it, combining AI-powered cash flow analysis with automated document verification and real-time credit scoring will capture a disproportionate share of the SMB market. Those that maintain manual processes designed for million-dollar commercial credits will continue to wonder why their small business portfolios are shrinking.

In my consulting practice, I see this dynamic playing out across the country. The banks that are winning in small business lending are the ones that have built separate digital origination channels specifically designed for smaller credits with automated underwriting below certain thresholds, streamlined documentation requirements, and decisioning timelines measured in hours rather than weeks. They are not abandoning relationship banking. They are using technology to handle the routine so that their relationship managers and lenders can focus on the complex transactions where human judgment and local knowledge genuinely add value.

Credit Risk in a Changing Landscape

None of this transformation happens in a vacuum. S&P Global forecast credit losses of $655 billion in 2026, a 7.5 percent increase from 2025, driven by commercial real estate stress, credit card normalization, and the lingering effects of rapid rate increases. Deloitte's outlook shows commercial and industrial loan demand recovering after a 5.6 percent decline in the first half of 2025, but with credit quality metrics that warrant close monitoring. Experian reports that 87 percent of financial institutions expect regulatory convergence of credit, fraud, and compliance functions which is a recognition that these risk domains are increasingly inseparable.

That convergence is one of the most consequential trends in lending today. A fraudulent loan application is simultaneously a credit risk, a fraud event, and a potential compliance violation. The institutions that maintain separate silos for credit risk, fraud prevention, and compliance monitoring are duplicating effort, creating gaps, and missing the connections between these interrelated risk categories. AI makes convergence possible, a single analytical platform that scores credit risk, flags fraud indicators, and monitors compliance obligations in real time but only if the organizational structure supports it.

The OCC and FDIC's December 2025 rescission of the interagency leveraged lending guidance signals a more principles-based approach to commercial credit supervision, and the agencies' proposed rescission of the 2023 CRA final rule in favor of the 1995 framework suggests a regulatory environment that is shifting toward simplification and reduced burden. Banks should welcome regulatory clarity while maintaining the credit discipline that these rules were designed to encourage. Deregulation is not an invitation to abandon sound practice, it is an opportunity to redirect compliance resources toward the emerging risks that actually threaten portfolio quality.

The transformation of lending is not a single initiative. It is the convergence of AI-powered decisioning, embedded distribution, private credit competition, SMB opportunity, and evolving credit risk dynamics into a landscape that rewards speed, intelligence, and discipline in equal measures. The institutions that will lead in lending over the next three to five years are the ones that treat technology and governance as complementary investments, not competing budget priorities. In the next article in this series, I will examine the cybersecurity and fraud challenges that are inseparable from this lending transformation because every loan originated through a digital channel is also an attack surface that must be defended.

Sources

Accenture, Banking Tech Trends Report 2026 (AI-first credit decisioning findings)

EY, Reconstructing the Financial Paradigm Through Intelligent Agents (lending case studies)

Experian, Global Insights 2026: 7 Shifts (composite AI adoption, credit-fraud-compliance convergence)

Finastra, State of the Nation Report 2026 (AI lending acceleration survey)

Capgemini, World CIB Report 2026 (private credit growth, $1.7T global assets)

11:FS, Making Embedded Banking Work (embedded lending conversion data)

BAI/ProSight, 2026 Banking Outlook Executive Report (SMB lending opportunity)

S&P Global, Banks Outlook 2026 (credit loss projections, $655B)

Deloitte, 2026 Banking & Capital Markets Outlook (C&I lending recovery, credit quality)

CSI/CITE Research, 2026 Banking Priorities Executive Report

CFPB, Winter 2025 Supervisory Highlights: Advanced Technologies Special Edition

CFPB Guidance on Credit Denials by Lenders Using Artificial Intelligence (adverse action requirements)

Interagency Guidance on Model Risk Management (SR 11-7); OCC Bulletin 2025-26

OCC Bulletin 2025-44, Rescission of Interagency Leveraged Lending Guidance (December 2025)

Interagency Guidance on Third-Party Relationships: Risk Management (2023)

Next in this series: Cybersecurity, Fraud, and the AI Arms Race

CCG Catalyst Consulting is a banking and fintech advisory firm that has guided over 600 financial institutions through core modernization, digital transformation, AI strategy, payments, contract negotiations, and M&A. Through its Bankers Fintech Council, CCG Catalyst also bridges the gap between banks and fintechs to accelerate responsible innovation. Managing Partner Paul Schaus is a recognized Top 25 Financial Services Consultant, and subject matter expert in banking, bringing experience across every level of the industry to the firm's advisory practice. Learn more at www.ccgcatalyst.com