Trust fears stall AI projects at medium & large firms

Gong's research found that 58% of medium and large businesses have stalled AI projects, with a trust gap cited as the main reason.

The findings were based on a survey of 2,056 business leaders in the UK and US, alongside Gong Labs' analysis of more than 25 million sales interactions processed on its platform.

Among UK respondents, 52% said AI projects had stalled, compared with 63% in the US. On average, 46% of planned AI investment had been paused - 47% in the UK and 44% in the US.

The data suggests trust concerns now rank above regulatory uncertainty in determining whether businesses move ahead with AI spending. Data privacy and security were cited by 34% of respondents as the main barrier to adoption, followed by explainability at 30% and model transparency at 28%. Regulatory uncertainty followed at 27%.

Businesses also expressed concern about falling behind. Three-quarters of respondents said their organisations were not getting enough value from AI, including 70% in the UK and 80% in the US.

Trust concerns

Gong Labs' analysis of sales calls pointed to the same pattern. One in four calls referenced security, while uncertainty around training data and how AI systems learn emerged as the most commonly discussed privacy and security issues.

The findings suggest buyers are focusing less on broad enthusiasm for AI and more on whether vendors can explain their systems, protect data and define clear limits on use. Respondents identified explainability as the leading assurance that would help them adopt AI tools with confidence, cited by 26% overall and 27% in the UK.

The ability to explain guardrails for protecting data followed at 25%. Security guarantees built into products and third-party audits or certification were each cited by 23%. A further 22% said transparency over training data use and model logic would help build confidence.

This creates a more demanding sales environment for software suppliers, as businesses scrutinise how AI tools generate outputs and whether sensitive information can be kept under control. In regulated sectors, those questions can determine whether projects move from trials to wider deployment.

Chris Peake, chief trust officer at Gong, said the issue has moved beyond compliance teams and into commercial decision-making.

"Security and AI trust are no longer back-office conversations; they are revenue conversations," Peake said. "Gong's research found that trust can be a performance multiplier when applied as part of an AI strategy. The competitive advantage delivered by AI is no longer up for debate, but the trust barrier remains for those using tools that have yet to establish this trust. By embedding enterprise-grade governance directly into the Revenue AI OS, we're helping the world's most successful teams bypass the doubt and move straight to acceleration."

Buyer demands

The research adds to the wider debate over how quickly businesses are willing to adopt AI tools when governance standards remain uneven. The figures suggest many companies have moved beyond initial experimentation and are now asking more detailed questions about auditability, security design and how models generate answers.

While the survey focused on medium and large businesses, the UK-US split points to a similar pattern in both markets. UK respondents were slightly less likely than their US peers to say projects had stalled, but concern levels were broadly aligned on explainability, model transparency and regulatory uncertainty.

For suppliers trying to turn AI interest into contracts, the results may be significant. If nearly half of planned spending is being paused, the barrier is not simply whether companies want to use AI, but whether they trust providers enough to proceed.

The survey was conducted by Censuswide among business leaders at medium and large organisations in the UK and US. Gong Labs' separate analysis examined aggregated, de-identified metadata and topic-level signals from sales interactions over the course of a year.