UK leaders overestimate everyday use of AI at work

Multiverse research has identified a gap between how UK leaders view workplace use of artificial intelligence and how employees say they use it day to day. Daily collaboration rates appear lower on the frontline than senior teams assume.

In the survey, 59% of leaders said AI had become part of everyday work. Among employees, 42% reported using AI daily-a 17 percentage point difference between perception and reported practice.

The findings suggest a visibility problem for executive teams as companies debate how quickly AI is spreading across offices and professional services. Adoption also appears uneven across roles and grades, with junior workers less likely to use AI tools than managers.

Autonomy gap

The widest difference between leaders and employees was on delegating entire tasks to AI. About 23% of CEOs believed employees were already handing over full tasks to AI systems, but only 8% of employees said they were doing so.

These figures point to a second disconnect beyond frequency: how much responsibility staff are willing to give AI tools. They also suggest leaders may be overestimating how mature AI use is in day-to-day workflows.

Seniority divide

Seniority accounted for a 30 percentage point gap in adoption: 52% of mid-level workers said they collaborated with AI daily, compared with 21% of junior employees.

A similar pattern emerged between managers and individual contributors. Nearly half of middle managers (48%) reported day-to-day AI use, versus 20% of individual contributors.

The concentration of adoption among managers could shape how AI projects are rolled out and governed. It also suggests employees closest to day-to-day delivery may not be getting the same exposure to tools, training, or expectations as management layers.

Task types

An addendum broke down gaps across several types of work, with leaders reporting higher use than employees in every category measured.

For automating repetitive tasks, 60% of leaders believed employees used AI for this purpose, compared with 36% of employees. For data analysis to support decision-making, the figures were 58% versus 33%. Optimising multi-stage processes was 52% versus 34%. Performing simple admin tasks showed a smaller but still meaningful difference: 52% versus 37%.

The spread across categories suggests the perception gap is not limited to a single department or use case. It covers routine work, analysis, and process activity-areas many companies cite as early targets for AI.

Training questions

The research also examined training among leaders. More than half (55%) said they had received fewer than five hours of formal AI training from their organisations. Another 58% said they relied on informal experimentation with tools such as ChatGPT to learn the basics.

Respondents also cited organisational and cultural barriers. Resistance to change was named as a primary challenge by 53% of leaders and 57% of employees. Around half of both groups pointed to a negative mindset toward AI as an obstacle to progress.

Even so, there was broad agreement that more training is needed. Some 85% of leaders and 78% of employees said more frequent training was essential to keep pace with the rate of change.

Multiverse said the findings point to a need for role-specific approaches, as adoption varies by seniority and day-to-day work and shapes how staff interact with AI tools.

Gary Eimerman, Multiverse's Chief Learning Officer, said the variation in use should inform how businesses plan skills development.

"AI is not a monolithic tool, and its application varies wildly between a junior developer, a middle manager, and a CEO. The 30% gap in adoption we see between seniority levels is a clear signal that the one-size-fits-all approach to AI is failing. To bridge this divide, businesses must move beyond generic training and implement custom AI upskilling paths tailored to the unique daily workflows of every individual."

The survey was conducted by Coleman Parkes and included 810 tech leaders and 1,190 employees. Respondents were based in the UK and the US, using a mixed method that included online and phone-to-web approaches.

Multiverse operates an upskilling platform focused on AI and technology adoption and works with more than 1,500 companies.