Is AI really helping your SMB? Study finds a quarter of execs can’t explain what their AI actually does | usagoldmines.com

• Study finds 25% of business leaders struggle to explain AI-generated outputs to stakeholders
• SMEs are trusting AI with complex financial tasks including audits and compliance
• Customers and investors are increasingly avoiding businesses that use unverified AI

Business leaders are being outfoxed by AI, with a quarter of those surveyed apparently unable to explain AI generated outputs, 12% “with a lot of difficulty” and only 22% “easily.”

Even more worryingly, the survey, by Startup.co.uk, appears to reveal a culture of “leave it to AI” regardless of the consequences.

The survey has also uncovered other concerns, from questions over GDPR to how SMEs are trusting AI to carry out sensitive financial work, including audits, expenses, and accounts payable automation.

Would you let your AI do this?

With GDPR responsibilities hovering over startups and dynamic new SMBs, the overuse of AI solutions could prove to be devastating if customer data is found to have been misused. Meanwhile concerns over the use of the technology could be affecting businesses that fail to verify AI work.

While the inability of startup leaders to explain AI generated work might be a surprise, the depth of AI use across financial tasks is of particular concern. The report revealed how 85% of small businesses are using AI to complete financial tasks of some sensitivity.

Among the figures are 37% using AI to automate accounts payable processes, 32% to handle audit and compliance, and 31% to manage spend and expenses. While there is some suitability for AI with fraud detection (26%) and performance insights (27%) with its capacity to analyze data in bulk, the inability of leaders to deal with questions about AI is a matter of concern.

Blind trust in AI

What the figures seem to indicate is an over-reliance on AI.

Zohra Huda, editor of Startups.co.uk, noted the data “highlights the bizarre corporate milestone we’ve reached in 2026. Founders are letting AI manage their fraud detection and accounting, but if a stakeholder asks how the numbers were calculated, a quarter of them can’t answer.”

The solution to that might be a bit of rehearsal, but it doesn’t change how data is being used by AI and what the implications for this are, and the likelihood of breaching Article 15 of the UK GDPR (concerning how data is processed).

“Blindly trusting a tech “black box” with sensitive financial data is a massive legal and compliance gamble,” added Huda. The takeaway for business leaders? If you can’t explain your AI’s logic to an investor or a customer, you’re potentially risking a very expensive GDPR fine.”

With the ICO able to fine businesses up to £17.5 million or 4% of global turnover for verified GDPR breaches, it is clear that startups relying heavily on AI need to up their game.