Meet the project - Acting on AI: Exploring Digital Constitutionalism in the Age of Artificial Intelligence

Artificial intelligence is no longer just a technical innovation – it is reshaping how societies legislate, govern, and define fundamental rights. Leading the way in asking what this means for democracy is Professor Kjetil Rommetveit, group leader for the CAS project Acting on AI: Digital Constitutionalism in the Age of Artificial Intelligence.

Rommetveit is Professor at the Centre for the Study of the Sciences and the Humanities (SVT) at the University of Bergen, where his research focuses on “science’s and technology’s public and political roles.” With Acting on AI, he and his team are examining how AI regulation transforms not only law and politics but also society’s very understanding of rights and values.

A Framework for Understanding AI and Politics

The project takes as its starting point the EU’s landmark AI Act – the first comprehensive attempt to regulate artificial intelligence across member states. For Rommetveit, this legislation opens up crucial questions about how technology, law, and politics intersect.

“What happens when techniques of risk management are introduced to the governance of legal rights? Can one meaningfully assess the ‘risks to a right,’ such as privacy? These are the kinds of questions the AI Act forces us to ask.”

Rather than only assessing the law’s technical details, the team investigates the broader forces shaping AI governance. They focus on three central dynamics: how imagined technological futures influence regulation and politics; the growing reliance on risk-based approaches; and the increasing use of technical standards as tools of digital governance.

By combining perspectives from science and technology studies (STS) with insights from law, political science, and technology, the researchers aim not to provide definitive answers, but to develop a conceptual framework that can inform and guide future debates on AI and democracy.

From Early Digital Governance to Today’s AI Act

Rommetveit has followed regulatory and technological shifts for more than a decade. “Together with colleagues (some of whom will join me at CAS) I have investigated the just mentioned regulatory and technological developments in digital governance and innovation since the early 2010s,” he notes.

What makes their approach distinctive is the theoretical lens they bring to these developments. “Our approach has been quite unique, in bringing fairly well-known perspectives from STS (and also law and philosophy) to bear on regulatory innovations and instruments (futures, risk and standards). These have not been so much studied from these theoretical perspectives.”

With the AI Act putting risk management and standardisation centre stage, the group’s earlier work takes on new significance. As Rommetveit puts it: “A concept of ‘digital constitutionalism’ can help here, although I emphasise that it is meant mainly as a problematisation: a way of opening up the relations between technology and politics to critical scrutiny, and also of questioning predominant simplistic notions of ‘artificial intelligence.’”

Interdisciplinary Collaborations

The project brings together perspectives from politics, technology, and law. “This is possible due to shared understanding based in the interdisciplinary field of science and technology studies, and has been consolidated in long-standing collaborations between partners,” Rommetveit explains.

Workshops will form the backbone of the project, each devoted to one of its three core themes: futures, risk, and standards. Each meeting will gather 10–15 participants, culminating in a final conference on AI and digital constitutionalism. Rommetveit also hopes to organise a session on AI and environmental issues.

Why CAS – and the Value of Basic Research

CAS offers an ideal setting for such wide-ranging work. “CAS is unique in providing space and opportunity for curiosity-driven and critically-minded research,” Rommetveit says. “During our stay at CAS my colleagues and I hope to be able to bring this (critical) approach to bear on AI and politics, and to create a coherent approach that can be used to study interactions of digital innovation, regulation and politics.”

He also situates the project within a broader defence of academic freedom and independence. “I do not believe in absolutely ‘pure’ and dis-interested knowledge. I do believe, however, in research that strives to be independent and that can ground questions, criticism and creative suggestions, directed at dominant policies and innovation agendas.”

At its best, Rommetveit argues, such research can contribute to new democratic and sustainable pathways: “Academics cannot do this alone, but we can contribute. We can collaborate with societal actors and provide legitimacy and importance to actors and perspectives that are otherwise ignored (for instance activists and publics, or neglected areas of research and innovation).”

He also welcomes the personal respite CAS offers: “Personally, I also look forward to some time off from administrative and teaching duties, to consolidate and renew my researcher networks, and create new avenues for publications and research projects.”

Looking Ahead

The project’s outcomes may not be final answers, but rather contributions to ongoing debates. “It is quite possible that this effort will fail as such. But I am confident that interesting things will come from it nevertheless,” Rommetveit concludes.

By situating AI regulation within the broader context of digital constitutionalism, the project not only contributes to academic debates but also addresses urgent societal questions about how rights and values evolve in a digital age.

By situating artificial intelligence within broader questions of democracy, law, and society, the project highlights that AI is never just about technology. It is also about values, politics, and the kind of future we collectively choose. In this sense, Acting on AI is not only a study of regulation, but an invitation to reflect on how we want to live with – and govern – intelligent machines. We look forward to following Professor Rommetveit and his team throughout this academic year as they explore these timely and pressing questions.