A new economic analysis published last week by the Federal Reserve Bank of San Francisco found something small business owners should sit up and read: the more aggressively a bank has adopted artificial intelligence, the fewer small business loans it tends to make.
The paper, released September 21, 2026, was authored by economists Mark Spiegel and Zheng Liu. It examined over 1,000 commercial banks covering more than 87% of total U.S. banking assets. The finding is counterintuitive. We’ve been told AI makes everything faster, better, more accessible. In lending, it’s doing the opposite — at least for Main Street.
What This Actually Means
Here’s the mechanism the SF Fed researchers identified: AI is exceptionally good at processing hard data — credit scores, tax returns, formal financial statements, payment histories. That’s the kind of clean, structured data large corporate borrowers generate in abundance. Small businesses, by contrast, tend to get loans based heavily on soft information — the banker who knows you personally, who understands the local market, who has watched you build something from nothing. AI can’t replicate that relationship. And because AI-heavy banks are optimizing for what the algorithm can evaluate, small business loans are getting squeezed out of the portfolio.
The researchers also found that high-AI banks earn a return on assets (ROA) about 0.38 percentage points higher than low-AI banks — so this isn’t bad banking, it’s profitable banking. The problem is that what’s profitable for the bank isn’t necessarily what’s available for you.
The Numbers Behind It
The SF Fed tracked AI adoption using job postings data from Lightcast, measuring the share of open roles requiring AI skills. By end of 2025:
- Large banks (over $100B in assets): 8.86% AI job share
- Mid-size banks ($10B-$100B): 4.48% AI job share
- Small banks (under $10B): 1.15% AI job share — and 84% of small banks had never posted a single AI-related job
The finding on SME lending: increased AI adoption is correlated with a declining share of small business loans (those under $1 million) in bank portfolios. The researchers note they cannot make fully causal claims, but the pattern held consistently across bank size categories.
Worth noting: the SF Fed economists acknowledged that small banks may use third-party AI vendors — which their job-postings methodology wouldn’t capture. So the gap between large and small bank AI adoption may be somewhat overstated. Still, the lending gap is real.
The Hustler’s Library Take
Here’s what the SF Fed paper doesn’t say outright but what every small business owner should take from it: your relationship with a local community bank or credit union is not a quaint throwback — it’s a strategic asset that is actively becoming more valuable.
Large banks are essentially telling small businesses: “We’ve automated the process, and the process says you’re hard to underwrite.” That’s a feature for their shareholders and a bug for your P&L. The businesses that get caught flat-footed are the ones who assume their longtime Chase or BofA branch will carry them through a tough quarter. The ones who win are already cultivating relationships at institutions that still rely on banker judgment — before they need a loan, not after.
There’s also a second angle here that’s being missed entirely: if AI is making big banks smarter at avoiding complex SME loans, it’s also creating a vacuum. Community banks and credit unions that lean into the relationship-lending model have a genuine competitive opening right now. If your lender is one of them, that relationship deserves more of your attention, not less.
For more on managing your business finances and hidden costs as a small business owner, that piece is worth revisiting with this new context in mind.
What You Should Do
1. Find out what tier your bank is in — now. If your primary banking relationship is with a large institution (assets over $100B), it’s worth understanding how AI-driven underwriting is changing their SME loan criteria. Ask your branch manager directly: “How does your current credit evaluation process work for businesses under $5M in revenue?” Their answer will tell you a lot about what the next credit cycle holds for you.
2. Build a relationship with a community bank or credit union before you need them. Open a business checking or savings account. Attend a local chamber event where their loan officers show up. Soft information works both ways: they need to know your story before a crisis, not during one. If you’re thinking about scaling your business, lender relationships are part of the foundation you need to build now.
3. Strengthen your hard data profile regardless. AI or not, every lender is looking at the same things: revenue trends, debt-service coverage, cash flow consistency. Make sure your books are current, your financial statements are clean, and your business credit profile is optimized. If any of that needs work, review the lessons from business owners who rebuilt — getting your financial house in order is always the first move.
4. Explore SBA-backed options proactively. SBA 7(a) and 504 loans come with a government guarantee that changes the risk calculation for lenders — including AI-optimized ones. If your growth plans include a capital raise in the next 12-18 months, talk to an SBA-approved lender now. The cost of capital planning should already be in your budget model.
Source: Federal Reserve Bank of San Francisco Economic Letter — “How AI Adoption Might Affect Bank Lending” (September 21, 2026). Additional context: SBA Loan Programs.
Want more intel like this in your inbox every morning? Join thousands of business owners who read Hustler’s Library daily. Join free here.
Ready to Know Where You Stand?
The Business Journey dashboard maps your exact position across all 13 stages. Track your progress, unlock resources for each step, and build with a framework used by thousands of founders at Hustler's Library.
No credit card required · Takes 3 minutes · Personalized to your stage
