Experiences with experimental research Lessons from a co-creative research project on AI effects in companies

Abstract

This policy brief outlines key lessons from experimental research on AI implementation in companies, emphasising complexities in methodology, co-creation, and legal-ethical considerations. AI adoption is rarely a straightforward process, requiring researchers to adapt flexibly to organisational changes and staggered rollouts. The co-creation approach, involving collaboration between researchers, companies, and policymakers, is essential but demands strong trust, clear communication, and continuous negotiation of different stakeholder interests. Methodologically, standard econometric tools often fail to capture AI-driven shifts in tasks and productivity, necessitating a combination of quantitative and qualitative approaches. Measuring AI’s impact is further complicated by the absence of clear “before and after” moments, gradual implementation strategies, and small affected sample sizes. Legal and ethical challenges, particularly regarding data protection and corporate sensitivities, require careful navigation. Despite these challenges, the study underscores the value of co-creative research in generating robust insights on AI’s role in the evolving workplace.
 

Download Policy Brief #9a 

Fregin, M.-C., Rounding, N., van der Velden, R., Levels, M., Özgül, P., & Steens, S. (2025). Experiences with experimental research Lessons from a co-creative research project on AI effects in companies. (pp. 1-9). ROA External Reports Vol. 09a.