SAIN Amsterdam
AI Safety in the Netherlands' largest city. Building a diverse community across UvA, VU Amsterdam, and the broader Amsterdam tech ecosystem.
Events
Past & upcoming events
Upcoming events: see all on lu.ma/AI_Safety_Ams.
Past events this academic year
Events: eventsams@safeainetherlands.org
Programs
What we run in Amsterdam
Courses
Technical AI Safety & Frontier AI Governance
We facilitate 2 BlueDot courses: "Technical AI Safety" and "Frontier AI Governance". We host both courses on-site in Amsterdam. In the previous iteration, we engaged 70+ individuals from varying backgrounds, such as students, PhDs, engineers, policymakers, and consultants. Facilitators range from PhDs, risk-management consultants, and an ELLIS assistant professor.
Technical Course
Focuses on the technical aspects of AI Safety, with extra sessions on mechanistic interpretability, adversarial attacks, complex systems, and more.
Governance Course
Prioritizes governance and policy aspects, dedicating time to case studies and real-world examples of regulatory, legal, and societal challenges.
Course Details
- Duration
- 6 weeks
- Workload
- 2h readings + 2h discussion / week
- Format
- On-site in Amsterdam
- Certification
- Certificate upon completion
- Selection
- Application-based
These courses are independently led by SAIN Amsterdam and are not affiliated with UvA/VU.
Questions about the course? eduams@safeainetherlands.org
Discussion group
Weekly discussion group
A focused group meeting weekly (~2 hours) to discuss, learn, and collaborate on AI Safety topics, guided by experienced mentors.
Technical AI Safety
Exploring technical approaches to AI alignment, interpretability, and safety evaluation. Reading and discussing cutting-edge research.
Questions about the discussion group? eduams@safeainetherlands.org
Research
Current research projects
A snapshot of current work. Research at SAIN Amsterdam isn't limited to the topics below. We're open to projects across mechanistic interpretability, alignment, evaluations, and AI governance. If you'd like to propose or join a project, see the Research Hub.
Investigating failure modes of Multi-Agent Debate
About
AI Safety at the heart of Dutch AI
SAIN Amsterdam operates in the Netherlands' largest city. The chapter engages a uniquely diverse community: from BSc students to professionals at companies like Deloitte and Shell, and from independent researchers to ELLIS assistant professors.
Formerly AISA (AI Safety Amsterdam), the chapter has rapidly grown into one of SAIN's most active branches, delivering courses to 70+ participants, presenting at the AI020 Conference and TEDxUniversiteit van Amsterdam, and building connections across the Amsterdam AI ecosystem.
The chapter is co-directed by Ana Paula Castillo Rodriguez and Andreea Chivu, with a structured team covering Research, Education, Events, and PR.
SAIN Amsterdam Team
Andreea Ioana Chivu
Co-Director
Ana Paula Castillo Rodriguez
Co-Director
Luan Fletcher
Discussion Group Lead & AI Technical Facilitator
Patrik Barták
Discussion Group Lead & AI Technical Facilitator
Ana Lucic
AI Technical Facilitator
Leon Eshuijs
AI Technical Facilitator
Lukasz Karwacki
AI Technical Facilitator
Satchit Chatterji
AI Technical Facilitator
Vlad Chibulcutean
AI Technical Facilitator
Jesselit Jimenez
AI Governance Facilitator
Maria Francesca Burger
AI Governance Facilitator
Ritu Raj
AI Governance Facilitator
Lot van Roosmalen
Events
Diego van der Mast
Public Relations
Maria Chiara Bargelli
Public Relations
Mark Mitrani
Public Relations
Maya Ozbayoglu
Public Relations
Thierry Blankenstein
Public Relations
Vin Sixsmith
Public Relations
Henning Bartsch
Advisory Board
Leonard Bereska
Advisory Board
Zoe Tzifa-Kratira
Advisory Board
Join & contact
Join our WhatsApp community to get involved. Subscribe to the national Substack for articles and updates across SAIN.
Email the right team to contact SAIN Amsterdam