AI Safety Atlas

Skip to main content

Textbook Impact

Textbook
1. Capabilities
2. Risks
3. Strategies
4. Governance
5. Evaluations
6. Misspecification
7. Misgeneralization
8. Scalable Oversight
9. Interpretability

AI SAFETY ATLAS

A comprehensive guide to AI safety and alignment research.

Leading research institute advancing AI safety and alignment in Europe.

Funders

Open Philanthropy

Cite this work as

Markov Grey and Charbel-Raphaël Segerie et al. 2025. AI Safety Atlas. French Center for AI Safety (CeSIA). URL: ai-safety-atlas.com

Get in Touch

Have questions about AI safety or want to contribute? Send us a message. This is an open source project. We are always looking for contributors or course collaborations.

Copyright: 2025 AI Safety Atlas • Text Content: CC BY-SA 4.0 • Code: MIT

Analytics: We use privacy-focused Plausible instead of Google Analytics. No cookies, fully GDPR/CCPA compliant. :)