Only one in ten UK organisations has successfully scaled AI or embedded it into core operations, according to new research from Accenture, leaving productivity gains concentrated at an individual level rather than being spread across enterprises.

The consulting giant’s Generating Impact report found a widening gap between AI use among workers and the organisational transformation. 

For instance, while nearly a fifth of employees now use gen AI tools daily and report higher‑quality outputs and faster delivery, with numbers tripling since 2024, only around 25% of workers said a major process in their team has been restructured around AI.

Meanwhile, more than a quarter of executives admitted that pulling the plug on AI would have “little immediate impact” on how their operation, suggesting the technology is still being applied at the margins. 

The slow adoption of AI in core processes has resulted in slow ROI, with almost half of executives (46%) reporting that AI has so far delivered little impact on profit or loss, underlining a challenge many firms are facing in moving from experimentation to scale.

Accenture said that the data show AI’s long‑term economic value will come primarily from higher‑quality outcomes and revenue growth, not cost efficiencies alone, provided organisations redesign how work gets done at scale.

For example, with the emergence of agentic AI, the study found nearly 82% of UK working hours could be enhanced by AI, up from 47% two years ago, but these gains are not yet visible in productivity data — a pattern seen in previous technology waves, where capability advanced faster than organisational redesign.

Accenture’s report also shows that the wider, long‑term economic impact of AI will be driven more by these higher‑quality outcomes, particularly as advanced agentic tools begin operating across multiple systems. 

Based on modelling across 17 industries, Accenture’s analysis shows the potential value created through revenue uplift is more than double the impact of cost savings and avoidance, which the firm highlights as an opportunity to reinvest capacity into areas such as faster product iteration, new offerings, and smarter capital allocation.

However, skills and delivery gaps need to be overcome to realise those benefits. Although more than half of UK workers (54%) have the appetite to reskill in response to AI, only 7% of executives said their workforce is fully prepared for agentic AI.

Moreover, 58% of execs said their organisation is not ready to integrate AI agents with core enterprise systems due to the complexity of their legacy IT estates.

With the UK’s Office for Budget Responsibility now factoring AI-driven productivity gains into its fiscal forecasts, these execution gaps could have severe implications for national competitiveness and public finances, problems the UK Government will be keen to address.

Recommended reading

• A Quarter of UK Firms Have Fallen Victim to AI Data Poisoning
• Want AI to Pay Off? Invest in Data and People, Gartner Says
• Only 2% of Firms Ready for Agentic AI Risks, New Report Warns

“AI’s productivity impact now sits at the heart of the UK’s economic resilience, not just business performance,” said Matt Prebble, head of Accenture in the UK & Ireland.

“While AI has officially joined the workforce, people are moving faster than their organisations. Personal productivity gains are visible, but unless workflows and operations are reinvented to scale AI, they can’t be translated at an organisational level.”

According to Accenture, unlocking AI’s economic potential requires coordinated reinvention across five core dimensions – strategy, work, workforce, digital core, and safety and security – with organisations acting across all five more than four times as likely to scale AI successfully as those focused on one area alone.