Workers Are Spending Nearly £1bn of Their Own Money on AI. What Does That Say About Leadership?

British workers are spending an estimated £958 million a year of their own money on generative AI tools they use for work.

My first reaction to Deloitte’s finding was not, “Look how quickly AI is being adopted.”

It was: what is happening inside organisations that makes employees feel they need to sort this out for themselves?

The figure makes a striking headline. Underneath it is a leadership question.

The workforce is getting on with it

Deloitte’s survey of 25,000 UK workers found that 63% knowingly use generative AI for work. Almost a quarter use it every day. One in six users pays personally for at least one tool they use for their job, and 31% use AI without their employer knowing.

That last figure raises legitimate questions about security and governance. But it also tells us something about people. They are experimenting, finding useful tools and, in some cases, investing their own money before their organisations have decided how AI should be used.

That looks remarkably like human agency to me. What are leaders doing with it?

The leadership gap

The finding that caught my attention most was this: 65% of GenAI users have not heard their leaders talk about AI with a clear understanding of it. Only 28% say their organisation has an AI policy they can find, understand and consider clear and current. Half have received no formal training.

Employees are already using the technology. Many are doing so without the direction they need to use it confidently and responsibly.

Every leader does not need to become an AI specialist. But leaders do need to understand enough to answer some fundamental questions:

• Why are we using AI and where do we want it to create value?
• What should our people use it for, and what should remain with them?
• Where must human judgement and authority sit?
• What happens to the time AI releases?
• What should our people become more capable of because AI is here?

These are leadership decisions. They cannot be settled by buying everyone a licence.

What happens to the time we save?

Workers in the survey reported saving an average of 70 minutes a week through GenAI. Deloitte found that most of the time saved is used to do more work for the same employer.

That deserves attention. Faster email can lead to more email. Faster analysis can produce more material for somebody else to read, check and respond to.

I see overload repeatedly in my work with senior leaders. Many do not need another way to do more. They need greater clarity about what deserves their finite attention.

I have started thinking about this as Attention Allocation. AI can increase the amount of information and analysis available to us. Human attention remains limited. Leaders need to decide where it matters most.

Using AI is different from redesigning work

Deloitte found that the most common workplace uses of GenAI are searching for information and drafting emails, each cited by 43% of users, followed by creating summaries at 31%.

These are useful tasks. They largely make familiar work faster, though. They do not necessarily change how the organisation works or what its people are able to do.

Deloitte’s separate 2026 Human Capital Trends research adds another concern. Among its UK respondents, only 18% thought organisations managed change effectively. More than three quarters said their wellbeing had suffered from too much change at work, while just 11% felt their organisation met their continuous learning needs.

If people experience constant change without clear direction or opportunities to learn, we should not be surprised when they devise their own solutions.

The question has changed

For several years, organisations have asked how to get their people to adopt AI. The survey suggests many people are already adopting it, paying for it and using it whether their employer knows or not.

The more demanding question is whether leadership can keep pace.

I think the next phase calls for Intelligence Allocation: deciding what work humans should do, what AI should do and where they should think together. It means identifying where human judgement creates the greatest value and protecting the experiences through which people develop it.

The £958 million is memorable. But at the next board meeting, I would be discussing why employees are willing to spend their own money on tools for work, why so many lack training and why the time AI releases so often becomes more work.

The workforce is already experimenting with the future of work. Leaders need to give that energy direction, purpose and room to develop human capability alongside the technology.

Sources: Deloitte UK GenAI Workforce Survey 2026 and UK findings from the 2026 Global Human Capital Trends report.