If you've wondered whether buyers are already quietly discounting businesses that haven't adopted AI, or paying a premium for the ones that have, you're not alone. AI's impact on business valuation is a reasonable thing to wonder about. It's also, right now, the wrong thing to lose sleep over.

There's the AI story you read about constantly: adoption is surging, productivity gains are real, and the technology is reshaping how large organizations operate. Then there's a separate, much quieter story: whether any of that is actually showing up in what a buyer will pay for a privately held business in the $2M–$50M range.

Those two stories get treated as one, and they aren't. The first is well documented and moving fast. The second, according to the people who are actually in the room negotiating these deals right now, is not moving nearly as fast. Confusing the two leads owners toward decisions, timing, positioning, even panic, that the evidence doesn't support.

If you've had an informal conversation with a PE firm in the last year, you've probably heard some version of an AI question already, even if it wasn't framed as a valuation issue. That's worth noticing. It means the conversation has already started on the buyer side, well before it shows up as a line item in an offer.

Start with the macro picture. Stanford's 2026 AI Index Report puts organizational AI adoption at 88% in 2025, with documented productivity gains of 14% to 26% in functions like customer support and software development (Stanford HAI, 2026). That's a real, substantial shift, and it's not in dispute.

Small businesses are a different story. NFIB's small-business survey found that only 24% of small employers used AI at all, and adoption scaled sharply with size: 21% for single-digit-employee firms versus 48% for firms with 50 or more employees (NFIB, 2025). Worth flagging: that survey was fielded in March 2025, so treat it as a snapshot from over a year ago, not a current read. Adoption at this size of business is likely higher today than it was then, and it's worth revisiting.

Now the number that actually answers the question you're asking. The IBBA/M&A Source Market Pulse Survey for the first quarter of 2026, which polls M&A advisors specifically on deals in the $0–$50M range, found that 67% report AI has had no material effect on the valuations they're seeing today. Only 12% see any current upside tied to AI capability, and just 3% see downside (IBBA/M&A Source, 2026). 15% said it's too early to tell.

The survey's own framing captures the moment well: AI is in the conversation, but it's not yet in the multiple.

Here's the part worth paying attention to. That 12%-upside, 3%-downside split isn't nothing, and the advisors surveyed expect the gap to widen as AI-capable and AI-lagging businesses start to diverge, particularly in automation-sensitive sectors. Diligence questions tend to show up well before that divergence reaches a price tag.

That's not a new pattern. It's the same sequence customer concentration and owner dependency followed before buyers started building them into offers directly (see Why Deals Fall Apart for how that plays out at the negotiating table). Buyers start asking. Then they start adjusting. The gap between those two moments is exactly where an informed owner has the advantage.

So expect questions about your AI tooling, your data practices, and how much of your process is still manual, and expect them before you see any of that reflected in a term sheet. An owner who has thought this through in advance answers those questions in minutes. An owner who hasn't spends weeks explaining something that should have taken minutes.

Think of it as the same discipline you'd apply to a change-of-control clause buried in a customer contract. It's not a problem until someone outside the business goes looking for it. The difference here is that you can see this question coming years in advance, which is more notice than most diligence surprises ever give you.

Not every business is equally exposed. The clearest, most defensible productivity gains in the Stanford data show up in customer support and software development, both process-heavy, tool-driven functions. Businesses built around relationships, judgment, and craft, the kind that make up a large share of this newsletter's readership, are further from where AI is currently proving itself.

That doesn't mean irrelevant. It means the businesses most likely to see AI show up in diligence questions first are the ones where a buyer can point to a specific, measurable process AI could plausibly touch. If that's your business, the questions arrive sooner. If it isn't, you still have some runway, but not forever.

None of this is a reason to chase an AI initiative for the sake of a future buyer. The evidence doesn't support urgency, and manufacturing a transformation story for a deal that may be years away tends to read as exactly that: manufactured. What it does support is an honest inventory.

Look at where your business actually uses AI today, if at all, using the same categories NFIB tracked: communications and documents, marketing, customer service, and back-office or accounting automation. Most small businesses that use AI at all are using it in the first two categories, not in core operations, so don't assume you're behind if you haven't automated the whole shop.

Document what you find, and document any process improvements as you make them, the same discipline you'd apply to explaining a margin swing or a cost-pass-through decision to a future buyer. Then revisit this question again within the next 6 to 12 months. Both the small-business adoption data and the valuation-impact reading are the kind of figures that move quickly, and what's true today may not be true by the time you're actually in a process.

The businesses that come out ahead here won't be the ones that reacted first. They'll be the ones that knew where they stood before anyone asked.