Who is more likely to use AI at work?

Nearly 50% of workers at the top end of the pay distribution reported using AI software, suggesting AI complements high-skill jobs.

AI use is highest in the finance and business services (32%) compared to other industries (17%) – and is lowest in public services like health, education, and public administration (15%).

Dr Henseke’s report looks at the evolution of work digitalisation since the 1990s, particularly the adoption of AI and robots across Britain and its impact on workforce reduction.

The report draws from the 2024 Skills and Employment Survey (SES2024), a nationally representative survey that aims to provide a rich picture of working Britain today. The survey is led by Cardiff University, in collaboration with UCL, Nuffield College, Oxford and the University of Surrey.

Through a mix of in-person and online surveys, adults aged 20-65 working in Britain were asked about their use of automated machinery and tools, AI algorithms and software at work.

The report found that in 2024, the digitalisation of work in Britain was nearly universal: 94% of the workforce used computers or automated equipment at their jobs; 80% said computers were essential to work.

This digitalisation provides a firm foundation for rapid AI adoption. Twenty-one percent of British workers reported using AI-powered software, which can include text generation, writing, conceptual reasoning and generative AI models like ChatGPT.

In between 2023 and mid-2024, AI was adopted rapidly, reflecting the growing use of software like ChatGPT. Within the same period however, there were no significant changes in robot use.

AI use tends to vary by education level, age and gender. Women were less likely than men to adopt AI, and employees in their 30s used AI more than those in their 50s. Those with a tertiary education were also more likely to use AI than secondary-school leavers – highlighting potential inequalities in access and utilisation.

The adoption of AI is also closely associated with workplace characteristics, varying by workplace size, management quality, work intensity, and scale of digitalisation.

Historically, introducing new technologies was associated with job cuts as business sought to reduce costs or enhance efficiency. However, this relationship vanished in the SES2024, suggesting that technology adoption does not inherently lead to job cuts – rather, the relationship between investments in technology and employee reductions are shaped by broader economic and labour market conditions.

As AI use continues to grow, the report advocates for better labour market intelligence to regularly assess its impact on work. It also calls for the development of policies to support lagging industries to trial ethical, high-quality AI solutions.

To ensure AI augments rather than replaces human labour, the report argues for employee engagement when developing AI strategies and emphasises the need for AI literacy training programmes – including pathways to retrain and upskill workers who hold at-risk jobs.

Dr Henseke says, “When we think of technology adoption over the last five years, it covers a great deal of change. The use of AI we’re observing with our data is in a very early, exploratory state.

“Moving forward, it is important to understand that the impact of new technologies on how we work, learn and live is constantly evolving, and subject to institutional decisions and policy.”