ProductNot yet confirmed elsewhere1 publisher3 min readPublished
67% of AI-using small and mid-sized firms run tools they built, Pax8 survey finds
Pax8's survey of 1,150 business leaders finds 67% of AI-using small and mid-sized firms have built an AI tool now in regular use. In every market most builders lack a formal, documented review, so governance, security and cost questions fall to the IT providers who serve them.
The Product Desk

Three US figures from Pax8's survey: 81% of AI-using businesses run solutions they built; 36% of builders have formal, documented review; 42% of AI-using businesses discovered shadow AI in the past six months.
| Figure | Value | Note | Claim |
|---|---|---|---|
| US AI-using businesses running solutions they built | 81% | 6 | |
| US builders with formal review | 36% | 4 | |
| US AI-using businesses that discovered shadow AI | 42% | in the past six months | 7 |
What happened
- Among builders, 52% are developing automated workflows that connect AI to other business systems, and 46% are building or testing AI agents that complete multi-step tasks on their own.
- The share of AI-using businesses running their own builds is 81% in the US, 76% in Sweden, 54% in the UK and 52% in Germany.
- Formal, documented review of in-house AI ranges from 23% of builders in the UK to 45% in Sweden.
- In the past six months, 35% of AI-using businesses discovered shadow AI, a term Pax8 defines as tools, workflows or agents running without official approval.
- 49% of businesses paying for AI over the past year were billed or charged for usage above what they expected, including 57% in the US and 65% in Sweden.
Why it matters
- exposure A client that shows an IT provider its AI-use policy has shown nothing about its builds, because Pax8 noted such policies do not cover the agents and workflows a business builds itself.
- cost Forecasting lands on whoever manages the account: usage-based pricing is the most-cited US payment route at 46%, and Pax8 said it is the hardest to forecast.
- decision IT providers have to decide whether inventory and review of client builds is a priced service or part of existing support, since most AI-using businesses in the sample are building.
Users spend the time AI saves checking the output. In Pax8's survey, 55% of decision-makers put in at least three hours a week of their own time going over, editing or fact-checking work produced by AI, and 7% put in ten or more [13]. The US figure is 68% [14]. Teams tell themselves users get the hours back. And 69% of AI-using businesses did say AI had freed up at least some capacity [15].
The 67% is a pooled figure. In the US, 81% of AI-using businesses run something they built; in Germany it is 52%, a gap of 29 points [6][21]. The UK sits at 54%, 2 points from Germany [23]. Pax8 put the margin of error at about 2.9 points for the full sample and 6.9 for Germany and Sweden, and said results for smaller subgroups should be read as directional [20]. A 2-point difference is inside both margins; 29 points is outside them. The sample also spans companies with 5 to 499 employees [19], roughly a 100-fold range in headcount [24].
In the US, 36% of builders have formal review and another 41% have review that is consistent but undocumented [4]. That leaves 23% in neither group [22].
The shadow AI counts need the same care. Sweden reports the most, 55%, and also the highest formal review rate [7][4]. The UK has the lowest review rate and the second-lowest shadow AI figure, 24% [4][7]. In our view, "discovered" counts what firms went looking for as much as what exists, so a high number can mean a firm is checking.
Pax8 sells AI and cloud software to small businesses through IT service providers and works with more than 47,000 of them [16]. Its report says the findings point to a growing role for partners [17]. Nick Heddy, Pax8's president and chief commerce officer, said in a statement: "The opportunity is significant, but so are the new demands around governance, security, cost, and accountability." [18] The survey also found seven in ten businesses willing to pay a premium of at least 6% for more control over providers' storage and use of their data [12]. That figure is a stated preference.
For an IT provider deciding what to offer clients tomorrow, we'd sort every client build with two questions: does it connect to another business system, and does it act without a person reading the output first? Both kinds are common, since 52% of builders make connected workflows and 46% are building or testing agents [5]. A tool that answers no to both needs a line in an inventory. A connected tool that a person still reads needs a check on what data it touches. An agent working alone needs a spending cap. A tool that is connected and acts alone needs documented review before the next invoice. The tradeoff is that the other three quadrants get a lighter touch than the formal review that reaches at most 45% of builders in any market Pax8 reports [4].
What to watch
- Whether the next quarterly Pax8 edition shows formal review of in-house builds rising above the current 23% to 45% range.
- Whether AI vendors offer the outcome-based pricing that 69% of respondents said would make them more likely to expand.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence45
- Adoption55
- Hype gap+15
- Incentives70
- Confidence50
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Pax8 surveyed 1,150 business leaders in the US, the UK, Germany and Sweden for its Q3 2026 SMB AI Pulse Report, published on Monday.
ReportedSupportedSource: Pax8 Q3 2026 SMB AI Pulse Report, as reported by The Next WebView cited source - [2]
Of the businesses that use AI, 67% have built at least one AI solution that is now in regular use, another 19% are prototyping, and only 13% rely solely on off-the-shelf tools.
- [3]
In every market surveyed, most businesses that build AI do so without a formal, documented review.
- [4]
Formal, documented review of AI built in-house ranges from 23% in the UK to 45% in Sweden; in the US, 36% of builders have formal review and another 41% use review that is consistent but undocumented.
- [5]
Among the builders, 52% are developing automated workflows that connect AI to other business systems and 46% are building or testing AI agents that complete multi-step tasks on their own.
- [6]
The share of businesses that use AI and run solutions they built is 81% in the US, 76% in Sweden, 54% in the UK and 52% in Germany.
- [7]
In the past six months, 35% of businesses that use AI discovered shadow AI, which Pax8 defines as tools, workflows or agents running without official approval; the rate was 55% in Sweden, 42% in the US, 24% in the UK and 19% in Germany.
- [8]
Some 40% of businesses have a documented policy on staff use of AI and 35% have informal guidelines; Pax8 noted that such policies do not cover the agents and workflows a business builds itself.
- [9]
Of the businesses that paid for AI in the past year, 49% got a bill or usage charge higher than they expected: 57% in the US, 65% in Sweden and 33% in Germany.
- [10]
US businesses pay for AI through usage-based pricing (46%), fixed subscriptions (41%) and per-seat pricing (40%); Pax8 said usage-based charges are the hardest to forecast.
- [11]
Some 69% of respondents said they would be more likely to expand AI with outcome-based pricing at a similar cost.
- [12]
Seven in ten businesses would pay at least 6% more for greater control over how providers store and use their data; in the US the figure was 81%.
- [13]
55% of decision-makers personally spend three or more hours a week reviewing, editing or fact-checking AI-generated work, and 7% spend ten hours or more.
- [14]
In the US, 68% of decision-makers spend at least three hours a week reviewing AI-generated work.
- [15]
Some 69% of businesses that use AI said it had freed up at least some capacity.
- [16]
Pax8 is a Denver-based marketplace that sells AI and cloud software to small businesses through IT service providers, and it works with more than 47,000 IT partners.
- [17]
Pax8's report says the findings point to a growing role for partners.
- [18]
"The opportunity is significant, but so are the new demands around governance, security, cost, and accountability," said Nick Heddy, Pax8's president and chief commerce officer, in a statement.
ReportedSupportedSource: Nick Heddy, Pax8 president and chief commerce officer, in a statementView cited source - [19]
Pax8 ran the survey online among owners, executives and managers involved in technology decisions at companies with 5 to 499 employees, polling 400 in the US, 350 in the UK, and 200 each in Germany and Sweden.
- [20]
The margin of error is about 2.9 percentage points for the full sample and 6.9 points for Germany and Sweden; Pax8 said results for smaller subgroups should be read as directional.
- [21]
The share of AI-using businesses running their own builds is 29 percentage points higher in the US than in Germany.
- [22]
In the US, 23% of builders fall in neither the formal-review group nor the consistent-but-undocumented group.
- [23]
The UK share of AI-using businesses running their own builds is 2 percentage points above Germany's.
- [24]
The survey's company-size range, 5 to 499 employees, spans roughly a 100-fold difference in headcount.
Sources
1 independent publisher whose own reporting we read for this story.
- thenextweb.comTwo-thirds of small firms using AI now build their own tools, Pax8 finds
1 article · October 13, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.