KPMG Quarterly Pulse: Making AI scale across the Enterprise

UK organisations are no longer asking whether to invest in AI, but how to make it work across the enterprise.

The research shows growing confidence in measuring AI’s impact on key areas of the business, with leaders certain about measuring ROI on productivity (76%), performance (71%) and decision making (67%).

But at the same time, organisations face ongoing barriers to scaling value, with 41% citing risk management – including cybersecurity and data privacy – as key AI challenges to AI strategy, followed by data quality and workforce readiness (32%).

UK businesses are moving, or planning to move, beyond early productivity tools to more advanced AI agents. 19% are actively scaling AI agents across multiple functions to automate and coordinate work end-to-end. However, with a third of organisations (34%) still in the exploration or pilot stage, the findings show that while the appetite and willingness to fully shift to AI agents is there, it's not cut and dried for UK businesses.

With AI agents comes additional levels of risk. Organisations are acutely aware of this, with 48% looking at AI agents as augmented support for employees and providing training for their workforce.

39% of UK organisations are adopting a ‘human-in-the-loop' approach where a human validates outputs but does not oversee each agentic action or decision. And 37% are going as far as not allowing AI agents to access sensitive data without human oversight.

While the picture may be mixed with regards to the maturity of scaling AI, doing so safely and effectively requires more than technology. It demands the right governance, controls and operating model from the outset.

As AI becomes embedded in day-to-day work, UK leaders must rethink their strategies to meet the needs of an AI-enabled workforce – and that includes both existing and future talent.

61% of UK businesses are currently upskilling or reskilling their current workforce or planning to do so. Just over half (52%), meanwhile, are recruiting new roles, such as prompt engineers or AI architects, and 48% are redesigning job roles.

But whether they are confident that their current pipeline can meet these needs is a different question. Overall, they are, with only 15% of respondents saying they are not.

The increased use of AI agents in the workplace will undoubtedly impact the employee workplace experience that businesses are looking for. Our survey reveals that UK businesses are now most focused on finding entry level employees who have adaptability and appetite for continuous learning (57%), critical thinking and problem-solving skills (52%) and technical or programming abilities (50%). And above all, they’re seeking out employees who have an ability to work effectively alongside the AI systems they have adopted and will look to adopt in the future.