What Centre pour la Sécurité de l'IA does
The Centre pour la Sécurité de l’IA (CeSIA) is a Paris-based, non-profit (association loi 1901) think-tank and policy research organization focused on preventing and mitigating major AI risks in France and Europe. It frames its mission around combining (1) policy analysis and public-policy recommendations for institutions, (2) education and public-facing awareness-building, and (3) technical AI safety research and engineering support in service of governance.
Operationally, CeSIA positions itself as an interlocutor for European and French decision-makers and emphasizes work products that connect safety research to governance needs—such as evaluation/benchmarking and enforcement-oriented policy arguments.
A distinctive part of CeSIA’s approach is that it publishes both technical and governance-focused outputs: it runs research and evaluation-oriented initiatives (e.g., the BELLS line of work for supervision/guardrails benchmarking) and simultaneously produces policy communications (including press releases and policy analyses), plus education assets such as its open AI safety textbook (“AI Safety Atlas”).