Britain’s AI danger zone: A third of UK firms show overconfidence in AI, report warns

A report that explores AI use and adoption in the UK has revealed an ‘AI danger zone’, with 32% of UK firms saying they completely trust AI, yet aren’t backing that trust with the governance, data controls or ethical safeguards needed to justify it.

The IDC Data and AI Impact Report: The Trust Imperative, commissioned by SAS, makes clear that responsible use of AI is fundamental to delivering business value, not a ‘nice to have’. While nearly a third of UK firms place complete trust in AI, only 8% back their confidence with the data governance, safeguards and ethical controls needed to make AI reliable and safe.

In short, the UK’s confidence in AI is uneven, and where it is high, it’s often not backed by a corresponding investment in ensuring AI is trustworthy. That mismatch affects a total of 44% of UK firms, which is defined in the report as the ‘AI trust dilemma’ (12% of firms have low trust, but high implementation of trustworthy AI - the opposite to the 32%).

When AI systems aren’t designed transparently or governed properly, organisations risk flawed outputs, regulatory breaches and erosion of customer trust. Conversely, firms that invest in responsible AI, from explainability and data privacy to bias prevention, are seeing higher productivity, better customer outcomes and stronger ROI according to the findings.

Speaking on the significance of the AI ’danger zone’ found within the report, Dr Iain Brown, Head of AI & Data Science at SAS Northern Europe said: “Overconfidence in AI is Britain’s blindspot. A third of firms are in a danger zone without realising it, blocking their ability to turn AI enthusiasm into competitive advantage and exposing them to risk.

“So, the message is clear, trust without governance isn't a workable AI strategy. To move from confidence to business value UK firms need to invest in robust data foundation, implement clear governance frameworks, build explainability and risk mitigation into model performance and underpin all of this with responsible AI policies. Only then will more UK firms progress their use of AI into the transformative stage and begin realising true business value.”

The UK business impact

According to the report the UK scores 2.93 out of 5 on the AI Impact Index (a quantitative measure that captures the realised business value and outcomes from AI investments), lagging behind the US (3.1), Spain (3.44), and Ireland’s impressive 4.06 rating. The report highlights several reasons behind this performance gap, beyond the trust dilemma.

This includes foundational data issues - UK organisations are more likely than global peers to cite data challenges such as poor data governance and data foundations not being centralised/optimised as limiting AI outcomes. The data also suggests hype-led adoption. Generative AI adoption is high at 78% but usage of more proven AI types lags behind at 58% - indicating a reliance on emerging technologies over more mature (and usually more reliable) solutions.

Also compared with Ireland, where 28% of firms report being in the ‘transformative’ stage of AI adoption, no counterparts in the UK believe they have reached this most advanced stage of AI adoption.

Dr Brown continues; “The absence of any UK firms at the ‘transformative’ stage of AI maturity is a warning sign that, while adoption is high, the UK is not yet realising meaningful, enterprise-wide value from AI.

“This stage represents the point where AI is fully embedded across the business, driving automation at scale, innovation and strategic growth. Without strong data foundations, governance and accountable deployment, UK organisations risk being outpaced by overseas rivals, where advanced AI maturity is already translating into higher productivity and ROI.”

With new regulatory frameworks on the horizon and competitors in Europe scaling up their AI capabilities, the findings represent a call to action for UK decision-makers: raise the bar on trustworthy AI or risk falling behind.

The IDC Data and AI Impact Report: The Trust Imperative draws on a global survey of 2,375 respondents conducted across North America, Latin America, Europe, the Middle East and Africa, and Asia Pacific. Participants included a balanced mix of IT professionals and line-of-business leaders, offering perspectives from both technology and business functions.