Artificial intelligence (AI) is increasingly being used in Danish workplaces. A peer-reviewed research article, co-authored by CEFAU researchers Michael Koch and Sarah Schroeder, shows that AI may affect the labor market in a fundamentally different way than previous technological advances. Unlike robots, AI does not primarily impact the labor market by replacing existing jobs or creating new ones. Instead, AI changes the composition of tasks within AI-exposed jobsand this can have large, heterogeneous effects on wages and skill requirements. If companiesand the broader economyare to reap the potential productivity gains from artificial intelligence, it requires continuous upskilling and training.

The key takeaways of this research brief are:

With a threefold increase over two years, the use of artificial intelligence is rapidly expanding among Danish companies
Robots and artificial intelligence affect very different groups in the labor market
Robots replace jobs; artificial intelligence changes job content
The results indicate a statistically significant positive effect on wages from increased exposure to artificial intelligence
AI's effect on wages is not uniform. Employees working in knowledge-intensive servicessuch as accountants, architects, or programmersexperience an average decline in wages, while lower-educated workers on average experience wage increases
AI is also associated with wage decreases for employees whose job content has changed significantly, but who have remained in the same position
ChatGPT has not led to a decline in the total number of jobs, but there has been a significant drop in employment among young people in professions highly exposed to artificial intelligence. In contrast, employment among workers over 50 has increased in AI-exposed professions following the launch of ChatGPT
The greatest risk of AI is not job lossbut a mismatch between tasks, skills, and wages
Education and continuous training in the use of AI are crucial if companies are to take advantage of the productivity improvements that artificial intelligence offers.

The research brief contextualizes and describes the findings in a Danish setting based on the article: Engberg, E., M. Koch, M. Lodefalk, and S. Schroeder (2025): Artificial intelligence, tasks, skills, and wages: Worker-level evidence from Germany. Research Policy, 54(8), 105285.

The research brief can be read here (in Danish).

CEFAU Brief: Proposal for an Alternative Benefit Indexation Mechanism that Ensures a More Stable Real Income for Recipients of Transfer Payments

Marie Møller Kjeldsen — Fri, 23 Jan 2026 16:19:42 +0100

In recent times, grocery prices have been a topic of debate. Many Danes experience that their everyday shopping has become more expensive which has led to political proposals for differentiated VAT and temporary food vouchers. Part of the background for these political proposals is that the real income of transferpayment recipients declined in connection with the surge in inflation in 2022 and has still not returned to its previous level. This policy note proposes an alternative indexation model under which the real income of transferpayment recipients would not experience prolonged declines the next time inflation jumps.

The analysis shows that:

In the event of sudden increases in inflation, there may be a prolonged decline in the real income of transferpayment recipients. This is due to the current indexation method. This has direct consequences for recipients of transfers, but also affects the overall economy as weaker private consumption can contribute to lower overall economic activity.
We propose a new indexation method that takes into account the expected development in inflation in the coming year. Current and wellfounded inflation forecasts are crucial to the methods ability to protect the real income of transferpayment recipients against unintended declines.
Had the proposed indexation method been in effect from 2020, transferpayment recipients would not have experienced as large or as prolonged a decline in real income during the inflation surge in 2022 as occurred under the current adjustment mechanism.
By basing indexation on expected inflation combined with a correction for previous forecast errors, it is possible to ensure that transfers develop over time in the same way as they do today. The proposed method does not change the total level of payments, only their timing.
Under certain reasonable assumptions, the proposed method can be implemented without requiring phasein periods or transitional arrangements.

The proposal can be read here.

RSS Feed

CEFAU Brief: AI and Its Impact on Jobs, Skill Requirements, and Wages

CEFAU Brief: Proposal for an Alternative Benefit Indexation Mechanism that Ensures a More Stable Real Income for Recipients of Transfer Payments