Sweden’s first AI Safety Day. A day to meet, share knowledge, and explore how we can build safer AI, together.
Organised by:
Premiss
Hosted at:
Lakefront, Emerge Stockholm Campus
Supported by:
Global Policy Research Group, Singapore AI Safety Hub, PauseAI Sweden, AI Safety Sweden, RegleraAI.nu, Simon Institute for Longterm Governance
Four Focus Areas
Governance
What a Swedish AI Safety Institute should do, and how labour-market and welfare institutions can prepare for a range of AI outcomes.
Technical
Strategic resilience, including continuity of access to foreign frontier models and compute, domestic fallback capabilities, and Nordic cooperation.
Talent Pipeline
What careers exist in the field and how we could turn Sweden’s talent into AI safety careers.
International Collaboration
Lessons from AI Safety Institutes and field-building organisations abroad.
Speakers & Panelists
Svante Linusson
Professor of mathematics and political candidate
KTH Royal Institute of Technology
Professor of discrete mathematics at KTH Royal Institute of Technology, doing research on combinatorics and teaching within Sweden’s national AI research initiative, WASP. His work also reaches beyond academia: he has helped improve Sweden’s electoral system, bringing mathematical rigour to the algorithms that determine how votes translate into parliamentary representation.
Olle Häggström
Professor
Chalmers University of Technology
Professor of mathematical statistics at Chalmers and member of the Royal Swedish Academy of Sciences. He has worked on AI safety for many years and is the author of books and articles on this topic.
Vivian Leung
Director of Programmes
Singapore AI Safety Hub (SASH)
Leads programmes and fieldbuilding at SASH, an independent, Singapore-based frontier AI governance and safety research organisation. SASH has launched the Singapore AI Safety Fellowship and the Frontier AI Security Training Technical Track and Policy Track.
Bahadir Sirin
Founder and President
Global Policy Research Group (GPRG)
Director of Policy Research at Accelerate Europe, where he advises European governments and institutions on AI policy across the full policy stack. He is also the founder and president of the Global Policy Research Group.
Jonathan Salter
EU Program Officer
ControlAI
EU Program Officer at ControlAI, working on AI governance. He has collaborated with CeSIA, Vinnova, the Odyssean Institute, the Talos Network and Effective Altruism Sweden on AI policy and field-building, and co-founded AI Safety Collab.
Johan Falk
AI Analyst
Falk AI
Has written several books on AI, regularly comments on AI news in the media, and publishes his own newsletter on major AI developments.
Henry Davidson
AI Policy Associate
Simon Institute for Longterm Governance
AI Policy Associate at the Simon Institute for Longterm Governance, where he leads work on international cooperation to make multi-agent AI systems safe and trustworthy. Previously, at the UK AI Security Institute, he managed research on human susceptibility to AI deception, manipulation and persuasion.
Ysaline Bourgine de Meder
Executive Director
Effective Altruism Sweden
Leads Effective Altruism Sweden and supports people who want to make a positive impact with their work, including by tackling AI risks. She also has policy experience in the Swedish government.
Axel Ahlqvist
AI Safety Researcher
Independent
Independent AI safety researcher. He recently co-authored a paper on making AI safety evaluations harder for models to distinguish from real-world use, in collaboration with Meridian Cambridge, the Cambridge Boston Alignment Initiative, the UK AI Security Institute and Anthropic.
Hunter Jay
Research Engineer
Australian AI Safety Institute
Research engineer at the Australian AI Safety Institute, focusing on compute verification and direct alignment research.
Program
A half-day of talks, matchmaking, and practical next steps, presented on one stage across four focus areas.
12:00 · Doors open
Doors open & coffee
13:00–13:10 · Welcome
Welcome Remarks
13:10–13:30 · Talk
Current AI Safety Incidents
Johan Falk
13:30–14:30 · Panel discussion
International Collaboration – Comparing Approaches Around the World
Svante Linusson, Vivian Leung, Jonathan Salter, Henry Davidson, Bahadir Sirin
14:30–15:15 · Talk
Where Humanity Stands
Olle Häggström
15:15–16:15 · Matchmaking
Matchmaking & fika
16:15–17:00 · Talk
AI Governance – How should Sweden prepare for an AI future?
Bahadir Sirin
17:00–17:45 · Talk
Technical AI Safety – Research on Eval Awareness
Axel Ahlqvist
17:45–18:30 · Talk
Talent Pipeline – Addressing risks from frontier models with your career
Ysaline Bourgine de Meder
18:30–20:00 · Networking
Networking & live music