Small businesses are cutting consultants, not compliance: the AI line banks have to read

Stax Payments surveyed 500 U.S.-based small and midsize business technology decision-makers in June 2026 and found that 80% have already used AI for work they would otherwise have paid a consultant or outside expert to perform ^[1][2]. The spending that AI is displacing first, in other words, is bought-in advice, which is exactly the ground banks and payment companies are trying to occupy with their own AI-paired services ^[14].

The interesting part is where adoption stops. Marketing and content creation was the most common use case at 61%, followed by graphic design at 49% and business planning at 45% ^[8]. Anything with a legal or compliance tail drops off a cliff: human resources at 25%, accounting and tax preparation at 25%, legal services at 22% and insurance at 16% ^[9]. That is a 45-point spread between the top use case and the bottom one ^[20], and every sensitive category sits at or below 25%, less than half the marketing figure ^[21]. Small firms are handing AI the work where a bad output costs an afternoon, and withholding the work where a bad output costs a filing.

Stax CEO John Cimba frames it as a shift "from whether small businesses should use AI to where they should use it" ^[11], and says the boundary is accountability rather than enthusiasm: "Businesses will use AI when it saves time, but also have boundaries around accuracy, compliance and accountability" ^[10]. He expects larger businesses to move into sensitive financial functions faster, because they are more likely to have technology staff managing AI strategy ^[12]. "Larger businesses have a bit more tolerance for learning," Cimba said. "Smaller businesses can't afford a mishap" ^[13].

Treat the headline enthusiasm numbers with some care. Nearly all respondents had already used AI in some form and reported a positive experience ^[3], which is a selected population, and the benefit figures come back almost uniformly high: 96% say AI speeds up decisions, 95% report more confidence in decision-making, 94% say it helps them serve customers better ^[4][5][6]. Stax is itself a small-business payment company selling into this segment ^[22]. Eighty percent expecting to increase AI use over the next six months is the more actionable number ^[7], because it is a forward commitment rather than a satisfaction score.

For banks, the strategic read is that the low-adoption categories are the ones attached to money movement and regulated data, and that is where incumbents still hold the accountability advantage. Andrew Jamison, CEO of small-business payment company Extend, told American Banker that agentic AI lets small firms get a more granular view of capital and cashflow with less need for financial management staff ^[16], and that what they want is a connection into accounting software to pull in data beyond the P&L ^[17]. Visa's global head of small business, Jonathan Kolozsvary, said small firms are constantly managing the gap between when they must pay and when they get paid, with distinct needs in payroll, supplier payments and working capital that consumer banking products do not address ^[18], and are looking for integrated experiences across payments, lending, deposits and cash management ^[19].

Watch whether the 22% legal and 25% accounting numbers move ^[9]. If they do, the vendor with the audit trail wins the SMB relationship; if they stay flat, banks are competing for marketing budgets they were never going to hold.