AI and law: “We need to think proactively”

Liane Colonna is an assistant professor in law and information technology at Stockholm University and director of the Swedish Law and Informatics Research Institute (IRI). She is doing research within the Wallenberg AI, Autonomous Systems and Software Program – Humanities and Society program (WASP-HS) and is interested in ethical and legal challenges in relationship to AI-driven practices in higher education as well as methodologically oriented research in the field of AI and Law. She has also worked as expert for the European Commission concerning AI and education. In addition, she has presented her research to UNESCO and at the United Nations Science Summit.

AI affects legal and societal governance systems

Within the research project “The Rule of AI – AI, Regulation, and Society” (funded by WASP-HS) Liane Colonna and her colleagues examine how AI affects legal and societal governance systems. The project asks how legal frameworks can both adapt to and shape the development of emerging technologies in ways that uphold fundamental rights and promote societal progress. One key aspect is to examine how regulation itself is changing: law is increasingly moving beyond traditional text-based rules toward forms that are embedded directly into technological systems and data-driven processes.

“Broadly, my work focuses on developing methodological approaches that help society to understand and analyze this shift. In particular, I explore how design-based regulation can be operationalized in practice, while ensuring transparency, accountability, and alignment with fundamental legal principles.”

Liane Colonna’s research explores key regulatory challenges including human oversight under the EU AI Act, the feasibility of implementing the GDPR’s right to be forgotten in machine learning systems, the allocation of responsibility across AI supply chains, and the effectiveness of AI governance in domains such as higher education.

Legal frameworks struggle to align with realities of AI

“A central finding across this body of work is that existing legal frameworks often struggle to align with the realities of AI system development and deployment, creating gaps related to oversight, accountability, and the distribution of legal responsibility. My research develops conceptual and methodological frameworks to better understand and operationalize these obligations, offering insights into how EU digital regulation – including the AI Act, GDPR, Cyber Resilience Act, and related instruments – can be interpreted and applied in ways that safeguard fundamental rights while remaining workable in practice,” says Liane Colonna.

You are also in charge of the research project AI and Law. What do you hope to achieve within this project?

“I aim to achieve three things. First, to develop a clearer theoretical understanding of how law is transforming in the context of AI, particularly the shift toward embedded and data-driven regulation. Second, to design methodological approaches to operationalize complex legal requirements. Third, to contribute meaningfully to ongoing policy debates about effective regulation.”

Essential with interdisciplinary projects

Liane Colonna believes that interdisciplinary projects are essential because the traditional tools in the lawyer’s toolbox are no longer sufficient to address the complexity of AI systems.

“Understanding and regulating emerging technologies requires insights not only from law, but also from engineering, social sciences, business, policy and beyond! These technologies shape – and are shaped by – technical design choices, economic incentives, and social practices. Bringing these perspectives together is therefore crucial to developing effective regulatory approaches that can adequately address challenges related to health, safety, the environment, and fundamental rights.”

Liane Colonna underlines that AI systems should not be designed solely for technical performance or operational efficiency.

“While optimization and functionality are important, technologies that affect people’s lives must also respect legal rules, fundamental rights, and broader societal values. This means considering issues such as data protection, information security, fairness, and accountability already at the design stage. A proactive approach is often far more effective than trying to address harms after they occur.”

Growing use of AI in education

The growing use of AI in education raises important legal, social, and ethical concerns, according to Liane Colonna. In particular, tensions can emerge between AI regulation such as the EU AI Act – and existing education regulation. These frictions become problematic when they risk undermining fundamental rights, including equal access to education, student privacy, and the quality of learning environments. At the same time, the introduction of AI systems may significantly reshape the role of teachers, potentially shifting their role from autonomous educators toward technology supervisors.

How can we handle this within the legal and political system?

“First, we have to build on the AI Act by developing clearer, sector-specific guidance for how its rules apply in practice. Second, strengthen co-regulatory tools like fundamental rights impact assessments by ensuring meaningful stakeholder involvement. Third, move toward a more distributed model of responsibility, rather than placing oversight solely on individual teachers.”

Is it possible to design new technology in compliance with the legal and ethical principles needed?

“Of course. We need to think proactively about law.”

What does it mean for you being a part of WASP-HS?

“Everything! The network has been truly inspiring. The opportunity to regularly engage with scholars from different disciplines has been one of the most valuable aspects of the experience. I have also been very generously supported and given the freedom to explore new ideas and directions in my research.”

Read more about Liane Colonna´s research
Read more about Wallenberg AI, Autonomous Systems and Software Program – Humanities and Society program (WASP-HS)

The Institute for Law and Informatics

The Institute for Law and Informatics (IRI) is an independent research unit within the Department of Law at Stockholm University, with a long-standing tradition of advancing scholarship at the intersection of law and technology. Established in 1968, IRI has developed into a highly project-oriented research environment. IRI has played a formative role in shaping the broader field of IT and law, both nationally and internationally. IRI is also deeply engaged in education. Law and Information Technology (“Rättsinformatik”) has been a compulsory component of legal education at Stockholm University since the early 1980s.
Read more about IRI Institutet för rättsinformatik

EU AI Act

The EU Artificial Intelligence Act, the most comprehensive legal framework that sets out harmonised rules for the development, placing on the market, and use of artificial intelligence in the EU entered into force on 1 August 2024. The AI Act has a gradual enforcement timeline.
Read about the EU AI Act