2nd International Conference

Large-Scale AI Risks

Ethics, Safety, and Governance
Date June 23–24, 2026
Location Leuven, Belgium
Venue Faculty Club, Grand Beguinage
Register
KU Leuven · Future of Life Institute
About

Shaping the Future
of AI Safety

The 2nd International Conference on Large-Scale AI Risks brings together researchers, policymakers, and practitioners from across disciplines to address the most pressing challenges of advanced AI systems. Hosted at the historic Faculty Club in Leuven's Grand Beguinage — a UNESCO World Heritage Site — LSAIR offers an exceptional setting for rigorous, interdisciplinary dialogue on the risks, ethics, and governance of transformative AI.

2
Days of Talks & Workshops
8
Research Topics
3
Keynote Speakers
175
Maximum Participants
Keynote Speakers

Leading Voices

Roman Yampolskiy

Roman Yampolskiy

University of Louisville

Computer scientist and AI safety researcher known for foundational work on AI controllability and impossibility results for safe artificial superintelligence.

Author of AI: Unexplainable, Unpredictable, Uncontrollable. His research focuses on behavioral biometrics, AI safety engineering, and the limits of controlling superintelligent systems.
Laura Weidinger

Laura Weidinger

Google DeepMind

Research scientist working on ethical and societal risks of large language models, with influential work on taxonomy of harms from AI systems.

Lead author of DeepMind's landmark papers on ethical and social risks of language models. Her taxonomy of LLM harms has become a foundational reference in AI safety research.
Werner Stengg

Werner Stengg

European Commission

Senior policy official and expert on EU AI policy and digital strategy, with deep experience shaping Europe's regulatory approach to artificial intelligence.

Key contributor to the EU AI Act and broader European digital strategy. Brings first-hand experience from the intersection of technology policy and democratic governance.
Programme

Research Topics

The conference covers eight interconnected areas addressing the large-scale challenges of AI. Click a topic to learn more.

01
Large-scale risks emerging from future AI systems
→
Investigating catastrophic and existential risks from advanced AI, including loss of control scenarios, misalignment at scale, and emergent dangerous capabilities.
02
Technical alignment plans for powerful systems
→
Approaches to ensuring advanced AI systems remain aligned with human values, including scalable oversight, interpretability, and robust goal specification.
03
Governance and control of agentic AI
→
Institutional, legal, and technical frameworks for governing autonomous AI agents, including accountability structures and control mechanisms.
04
Ethical considerations and frameworks for AI risks
→
Moral philosophy applied to large-scale AI risks, including questions of responsibility, precautionary principles, and intergenerational ethics.
05
Identification of present large-scale risks
→
Documenting and analyzing large-scale risks from currently deployed AI systems, including systemic biases, market concentration, and infrastructure dependencies.
06
Psychology, cognitive science and large-scale risks
→
How cognitive biases, risk perception, and group decision-making affect our ability to anticipate and respond to AI risks at scale.
07
Policy considerations to prevent catastrophic outcomes
→
Concrete policy proposals for national and international regulation, including compute governance, licensing regimes, and international coordination.
08
Philosophical considerations of large-scale AI risks
→
Foundational questions about consciousness, moral status, existential risk methodology, and the epistemology of reasoning about unprecedented threats.
Interdisciplinary

Structured Networking

LSAIR is designed to foster meaningful connections across disciplines. Our structured networking sessions ensure every participant can engage with researchers outside their immediate field.

Cross-Discipline

Curated Matching

Connect with researchers from philosophy, computer science, policy, cognitive science, and more through AI-powered introductions on Swapcard.

Facilitated

Structured Sessions

Dedicated networking slots in the schedule, with pre-arranged meetings and open discussion rounds for spontaneous connections.

International

Global Community

Join scholars, practitioners, and policymakers united in addressing AI's most critical challenges.

Venue & Travel

Join Us in Leuven

Faculty Club, Grand Beguinage

UNESCO World Heritage Site · Groot Begijnhof 14, 3000 Leuven, Belgium

The conference takes place at the Faculty Club, nestled within Leuven's magnificent Grand Beguinage — a 13th-century UNESCO World Heritage Site. This beautifully preserved complex of historic houses and gardens provides an inspiring and contemplative setting for scholarly exchange. Visit venue website →

Brussels Airport (Zaventem)

Direct train to Leuven in just 15 minutes. Trains depart every 15–20 minutes from the airport station located below the terminal.

Brussels South (Charleroi)

Take the shuttle bus to Bruxelles-Midi station (~1h, ~€13), then a direct train to Leuven (~25–30 minutes).

Important: Do not take trains to "Louvain-La-Neuve Université" — that is a different city. Make sure your ticket is for Leuven.
Accommodation → Visit Brussels →
Registration

Register Now

June 23–24, 2026
Faculty Club, Grand Beguinage · Leuven, Belgium

Conference registration covers all sessions and networking. The dinner package adds an evening of conversation at the historic Faculty Club.

Conference
€200
Conference + Dinner
€300

Payment Options

Invoice — Choose this if your institution or employer handles the payment. You will receive an invoice that can be forwarded to your finance department.

Receipt — Choose this if you are paying personally and need proof of payment for reimbursement. You will receive a reference number to make your payment directly, and a receipt after payment.

Register Now
Organizers

Organizing Committee

Torben Swoboda

KU Leuven

Lode Lauwaert

KU Leuven

Risto Uuk

Future of Life Institute

Organized by

Supported by the Chair of Ethics and AI, KU Leuven