A new industry report has found that although artificial intelligence is being adopted quickly across UK organisations, only a minority are successfully converting the insights they generate into meaningful business decisions.

The study, AI in Business: From Reporting to Actionable Insights, was produced by consultancy Differentis in collaboration with 4th-IR and Swiss-based AI education body LAC².

Drawing on interviews with technology strategists, AI specialists and transformation leaders, the research examines the widening gap between rising AI investment and the relatively limited business impact many organisations are achieving.

One of the central findings is that 72 percent of organisations now use AI to generate insights, yet fewer than a quarter turn those insights into specific actions. Although spending on data platforms has increased, the authors say cultural, operational and organisational barriers continue to slow progress.

The report also highlights an imbalance between technical output and practical understanding. Nearly half of those surveyed said their AI models produce results that teams either do not understand or do not trust. Only 18 percent have established feedback loops linking AI recommendations to real-world business performance.

A further challenge is the rush to deploy systems before organisations are ready. Respondents described fragmented pilots and projects launched at speed but without the structures needed to scale or deliver a return on investment. At the same time, 74 percent said they would prioritise transparency and auditability over rapid automation, with those focusing on explainability reporting higher adoption and stronger stakeholder confidence.

Dave King, Managing Director at Differentis, said the findings show that “AI is no longer about proof of concept. It is about proof of understanding.” He argued that the real challenge now lies in building the capabilities and confidence needed to act on AI-generated insights.

Thomas Gustinis, CEO of 4th-IR, said the next phase of AI adoption “is not technical; it is human”, noting that organisations making the most progress are those investing in clarity, trust and responsible use.

The report also points to a growing need for education and skills development, an area in which LAC² specialises, as companies struggle to embed AI into everyday operations.