A free in-person bootcamp in Switzerland introducing students and early-career researchers to AI safety through technical and conceptual coursework. The camp covers alignment, mechanistic interpretability, and governance tracks.
A free in-person bootcamp in Switzerland introducing students and early-career researchers to AI safety through technical and conceptual coursework. The camp covers alignment, mechanistic interpretability, and governance tracks.
People– no linked people
Updated 04/02/26Funding Details
Updated 04/02/26- Annual Budget
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- $51,248
Org Details
Updated 04/02/26The Swiss AI Safety Summer Camp is a free, in-person educational bootcamp designed to introduce students and early-career researchers to the field of AI safety. The inaugural edition ran September 4-16, 2023, at juhui Melchtal near Lucerne, Switzerland, hosting approximately 20 participants alongside a team of eight volunteer organizers and teachers. The camp offers a multidisciplinary curriculum spanning technical and conceptual AI safety. Technical tracks include deep reinforcement learning, convolutional neural networks and interpretability, transformer architecture, and GPU and compute governance. Conceptual tracks draw on AI safety fundamentals readings and explore the alignment problem from philosophical and strategic angles. Researchers give conferences on AI risks including Q&A sessions. Participants are expected to have a background in linear algebra, calculus, probability, Python, and basic deep learning, and the organizers provide preparation resources for those needing support. The camp explicitly prioritizes diversity, strongly encouraging applications from women, racialised people, people with disabilities, and LGBTQ+ people. All costs including housing, meals, and optional transportation assistance are covered for participants. The 2023 edition was funded by a CHF 45,165 (~$51,000) grant from Open Philanthropy. Similar camps with related content run in France, Germany, and Denmark.
Theory of Change
Updated 04/02/26The camp aims to reduce existential risk from AI by expanding the pipeline of researchers and practitioners working on AI safety. By providing free, high-quality in-person training to students and early-career researchers, it lowers barriers to entering the field and builds a community of safety-oriented practitioners. Participants gain both technical skills (interpretability, governance) and conceptual grounding in the alignment problem, making them more likely to pursue careers in AI safety.
Grants Received
Updated 04/02/26Projects– no linked projects
Updated 04/02/26Discussion
No comments yet. Be the first to share your thoughts.