Artificial Intelligence in Firms: A Synthesis of Micro-Level Evidence on Productivity, Work, and Skills

The latest ai:conomics policy brief brings together scientific evidence on how AI affects work, productivity, and skills within organisations.

The picture that emerges is clear: AI is no longer an abstract future scenario, but part of everyday business practice. Generative AI in particular is reshaping tasks, workflows, and decision-making. The effects are not uniform. AI can boost productivity and work quality, especially for less experienced workers, but outcomes depend heavily on how AI is embedded in work processes, governance structures, communication, and training.

A key conclusion is that AI currently changes work mainly through shifts in tasks and skill requirements, rather than through widespread job losses. At the same time, increasingly autonomous AI systems that act and decide independently demand sharper thinking about responsibility, human oversight, and the distribution of risks and benefits.

Successful AI implementation is not merely a technological challenge. Whether it works depends on how organisations approach it: leadership and work design on one side, employee participation, HR, and co-determination on the other. Those who take that seriously improve the chances that AI will genuinely work for the organisation and the people within it.

Download the ROA Policy Brief

Fregin, M.-C., Eijkenboom, D., Özgül - Persyn, P., Pardesi, M., & Rounding, N. (2026). Maastricht University, Research Centre for Education and the Labour Market. ROA Policy Brief No. 002E https://doi.org/10.26481/umarpb.2026002E