Research agenda

My research is organized around three workstreams. Drawing on economic sociology and science and technology studies, as well as the sociology of professions, I analyze how constantly changing and fast-moving AI systems are adopted and governed within public institutions, in markets, as well as how they interfere with professional judgements, expertise and working practices.

Public-sector AI governance

How do states orchestrate AI development, and how do public-private actor coalitions co-configure the social, technical and legal infrastructure of an AI-using society?

This workstream examines the social, technical and legal infrastructures through which technology and society are co-constituted. Nordic states increasingly take on the connective work of rolling out and scaling AI systems across the public sector. Together with Torben Elgaard Jensen, I conceptualize this work as infrastructuring and study it comparatively across Denmark, Sweden, Norway and Finland, based on a mapping of more than 300 public-sector AI initiatives and funding decisions.

I am also interested in how actors key to the infrastructuring of public sector AI articulate visions of a common good, and of what an AI society can become.

Algorithmic markets

How do algorithmic systems take root in some market contexts and not in others, and what does this mean for people’s life chances?

This workstream addresses uneven algorithmization, that is, the selective use of algorithmic technologies in some contexts but not in others. Contrary to expectations of a data imperative, Danish life insurers use machine learning and behavioral data to prevent long-term illness, yet refrain from using the same techniques to individualize prices. My work explains this pattern through the moral authority of data professionals and through the legitimacy alignments, or their absence, between the market and its political and public audiences. I focus on markets that shape people’s life chances, such as insurance and credit, and pay particular attention to moralized markets, where legitimation is a central principle of market exchange relationships.

Representative studies are Moral authority over risk classifications in Socio-Economic Review (2025) and From risk transfer to risk prevention, with J. O. Willers, in the Journal of Organizational Sociology (2026).

Professional work and ethics

How do professional cultures and ethical frameworks shape algorithmic practice, and how is AI reshaping professional expertise and judgement?

This workstream examines how professional cultures and ethical frameworks shape practice within algorithmic markets. My empirical focus is on data scientists, a new profession that operates in more fluid and less institutionalized ways than established professions such as actuaries. I study how data scientists gain authority in organizations and how competing notions of fairness inform the classification of risk. I also examine how ethical expertise becomes a source of professional power in EU AI governance, and how AI and generative AI alter professional expertise and the boundaries between professions.

Representative studies are Data is the new money in the Journal of Professions and Organization (forthcoming) and Solving AI ethics?, with J. O. Willers, in the Journal of European Public Policy (2025).

Methods

My projects combine ethnographic observation and interviews with document analysis. On the quantitative side, I use statistical modelling and network analysis as well as multiple correspondence analysis.