Aarhus University Seal

Presentation of preliminary findings on clinical AI at international conference at CBS

On April 9th, Kalle Kusk and Asbjørn Malte Pedersen presented preliminary findings from the research project at the Controversies of AI Society conference at Copenhagen Business School. The presentation focused on clinicians’ multiple roles, the structuring role of regulation in AI innovation, and tensions between different epistemologies in clinical AI projects.

The conference hall at CBS during Controversies of AI Society, where the research project presented preliminary findings on the implementation and scaling of clinical AI in Denmark.

On April 9th, the research project took part in the conference Controversies of AI Society, held at Copenhagen Business School (CBS), Denmark, on April 9-10th. The conference was organised by the research projects Algorithms, Data & Democracy (ADD) and Strategizing Communication and Artificial Intelligence (SCAI).

At the conference, Kalle Kusk and Asbjørn Malte Pedersen presented the paper “I didn’t become a physician to be a glorified secretary”: Controversies in clinical AI implementation and scaling. The abstract was authored by Claus Bossen, Kalle Kusk, and Asbjørn Malte Pedersen.

The presentation shared preliminary findings from the project’s broader work on the implementation and scaling of clinical AI in Denmark. So far, the project includes eight cases that examine different aspects of how clinical AI is developed, implemented, and put to use in healthcare settings.

The presentation highlighted three main themes. First, it addressed the clinician’s multiple roles in clinical AI projects. Clinicians do not only act as healthcare professionals, but also as researchers, project managers, and, at times, practical interpreters of legal and regulatory requirements. These roles are vital to the progress of clinical AI projects, yet they are also fragile and demand new competencies.

Second, the presentation showed how regulation and legal frameworks do not merely impose requirements on clinical AI systems. They also shape which models are developed and how clinicians are able to work with them in practice. Regulation should therefore be understood not only as a framework for innovation, but also as an active and structuring force within it.

Third, the presentation explored the tensions produced by two diverging epistemologies in clinical AI projects. On the one hand, there are the logics of AI innovation. On the other hand, there is the traditional medical approach to development, evaluation, and adoption. These differences create ongoing tensions in the implementation and scaling of clinical AI.

We are grateful for the opportunity to participate in the conference and to present the project’s preliminary findings.

You can learn more about the conference here. Our extended abstract is also available in the book of abstracts, published by Aalborg University Open Publishing.