Policy Briefing - AI and employee pay in the UK: new evidence, and implications

Artificial intelligence (AI) is rapidly becoming embedded in UK workplaces, impacting how work is organised, monitored and rewarded. While public debate has focused heavily on automation and job displacement, far less attention has been paid to how AI affects workers’ pay and pay inequalities in the workplace.

This briefing is based on the recent article “Artificial Intelligence Technologies and Employee Pay in the United Kingdom: Evidence From Matched Employer–Employee Data”, published in the British Journal of Industrial Relations by Felix Schulz, Danat Valizade, Mark Stuart, Magdalena Soffia and Jolene Skordis (Schulz et al., 2025). The article presents a research collaboration between the Institute for the Future of Work (IFOW) and the ESRC Research Centre for Digital Futures at Work (Digit), funded respectively by the Nuffield Foundation (Pissarides Review into the Future of Work and Wellbeing) and the Economic and Social Research Council.

Wider findings from Digit’s research and the Pissarides Review indicate that while AI adoption is not yet radical (only around a third of UK employers had invested in AI-enabled technologies in the past five years), these technologies are beginning to affect workers’ quality of life and job quality. Building on this early evidence, Schulz et al. (2025) examine the likely impacts of AI adoption on pay outcomes using matched data from over 2,000 UK employers and more than 5,400 workers.

The emerging findings are promising. AI adoption is associated with higher pay and the largest relative gains accrue to lower-qualified and lower-skilled workers. However, these gains are contingent on two critical factors:

workers’ frequent use of AI; and
employee involvement in decision-making. Where employees are consulted or involved in pay negotiations, AI-related pay benefits are more evenly distributed.

These findings are consistent with the principle of human-centred or pro-worker AI (Acemoglu 2025). We would argue that AI can support a more inclusive and equitable labour market, but to do so necessitates a supportive policy environment that actively shapes how AI is deployed and furthers skill development and employee participation in workplace governance. The alternative is a fragmented AI transition that benefits a narrow segment of the workforce while leaving structural inequalities intact.