About the kickoff event
The SIG-AI-ACT kickoff event, taking place on 6th March, marks the official launch of the initiative and brings together researchers, practitioners, and stakeholders interested in accountable, transparent, and trustworthy AI. The event will feature two keynote talks that set the vision and research agenda for SIG-AI-ACT, addressing both foundational challenges and emerging opportunities at the intersection of AI systems and regulatory frameworks such as the EU AI Act. Following the keynotes, participants will engage in breakout sessions, designed to foster interactive dialogue on concrete research directions, practical implications, and opportunities for cross-disciplinary collaboration.
Most importantly, we aim to build a community at the intersection of AI research, practice, and regulation and therefore seek topics and questions to be discussed in our breakout sessions. We kindly ask you to contribute such questions during registration.
For inspiration, example questions include:
- What evidence do you currently rely on to trust AI-based decision-support systems in your domain?
- What are the main technical, organizational, or legal barriers you face when integrating the EU AI Act into existing AI workflows?
- Which aspects of the AI Act are hardest to translate into concrete technical requirements or evaluation criteria?
- What best practices have you seen that effectively balance trustworthiness, transparency, and privacy in real deployments?
- How can regulation such as the EU AI Act enable innovation rather than hinder it, and what technical or organizational mechanisms are needed to support this?
Details
- Date: 6 March, 2026
- Time: 9:30–17:00
- Location: Aula Conference Center, TU Delft – Mekelweg 5, 2628 CD Delft (Room Frans van Hasseltzaal)
Keynote Speakers
Prof. Asia Biega — Max Planck Institute for Software Systems (MPI-SP), Germany
Talk Title: TBA
Abstract: TBA.
About: Asia Biega leads the Responsible Computing group at MPI-SP, where her research focuses on developing, analyzing, and computationally operationalizing principles of responsible computing, data protection and governance, and digital well-being.
Prof. Andrej Zwitter — University of Klagenfurt, Austria
Talk Title: Trust in the Age of General Purpose AI: Cybernetic Governance and the EU AI Act
Abstract: The European Union’s Artificial Intelligence Act marks a decisive shift toward proactive and risk-based governance of artificial intelligence, with General-Purpose AI (GPAI) emerging as a central regulatory challenge. GPAI systems are not confined to single use cases but operate as foundational infrastructures embedded across sectors, applications, and regulatory domains. This keynote examines how the AI Act’s GPAI framework both reflects and struggles with a broader structural problem: the convergence of digital technologies and the resulting convergence of governance regimes.
About: Andrej Zwitter is Professor of Humanities and Digitalization as well as Director of the Digital Age Research Center at the University of Klagenfurt, Austria. He served as Dean of the Faculty Campus Fryslan and founding dean of the University College Fryslan, University of Groningen, Netherlands. He was a founding board member of the WEF Chapter Zero, Netherlands and of the urban data centre Data Fryslân. His expertise covers data governance, law and ethics of digitalization, as well as emergency management and sustainabilty governance.
Program Summary
Welcome and keynote by Asia Biega
Parallel breakout groups (Algorithms, Industry, Policy) to map participant interests, current practices, concerns, and barriers to AI Act integration.
Informal discussions and networking among participants from academia, industry, and policy.
Keynote by Andrej Zwitter on legal and governance challenges
Solution-oriented breakout discussions to identify best practices, concrete community outputs, and future research directions.
Panel discussion and synthesis of breakout insights, followed by closing remarks and discussion of next steps for the SIG-AI-ACT community.
