When businesses learn there’s no grant to simply buy AI, the next practical question is how to finance it — and for many US small businesses, the answer is an SBA 7(a) loan. Here’s how it works for technology and AI in 2026, cited to the SBA. (dgm implements AI and builds the ROI case; the loan itself comes from an SBA lender.)
What a 7(a) loan is
The 7(a) loan is the SBA’s primary program: a bank-issued, SBA-guaranteed loan for small businesses. Its strength for AI is flexibility — the use-of-proceeds rules cover working capital, equipment, and technology, so software and AI investments fit within the standard program. There’s no special “tech 7(a)”; it’s the regular loan used for a technology purpose.
The honest framing: this is debt you repay, not a grant. But since no broad grant exists to buy AI, financing is frequently the most practical path — especially when the AI pays for itself.
How much you can borrow — and the July 2026 change
The individual 7(a) maximum is $5 million. The notable 2026 update: effective July 4, 2026, the SBA doubles the cumulative 7(a)+504 limit to $10 million, so a borrower can hold up to $5 million in 7(a) plus up to $5 million in 504 financing combined. The individual 7(a) cap itself stays at $5 million.
Terms and rates
Terms generally run up to about 10 years for equipment and working capital and up to 25 years for real estate. Rates are variable, typically structured as a base rate (such as prime) plus a spread that’s capped by loan size — and as of 2026 lenders may use alternative base rates in addition to prime. Because the exact maximum allowable rates change, confirm current figures with an SBA-approved lender rather than relying on third-party rate roundups.
7(a) vs 504 for technology
The 504 program is for major fixed assets — real estate and equipment — and generally does not cover software on its own. So for AI software and most technology spend, 7(a) is the relevant program; 504 comes into play if you’re financing AI-enabled machinery or facilities.
When financing makes sense
A loan is worth taking when the AI generates a clear, near-term return — cutting labor cost, accelerating revenue, or replacing more expensive tools. That’s why the ROI case matters: it both justifies the borrowing and strengthens your loan application. If instead you’re doing genuine AI R&D, look at SBIR/STTR/NSF grants and the R&D tax credit before taking on debt.
How dgm helps
dgm implements osFoundry and other AI for US businesses. We build the ROI case that makes a technology loan worth taking — and that lenders want to see — and then deliver the implementation. The 7(a) loan comes from an SBA-approved lender; the AI that justifies it comes from us.