https://cesia.org/fr/2026-03-28https://cesia.org/fr/our-work/2026-03-27https://cesia.org/fr/publications/2026-03-25https://cesia.org/fr/publications/bells-un-benchmark-pour-evaluer-la-fiabilite-des-systemes-de-supervisions-des-llms/2026-03-25https://cesia.org/fr/publications/communique-de-presse-du-cesia-sur-la-declaration-conjointe-france-chine-sur-lia/2026-03-25https://cesia.org/fr/publications/dans-quels-domaines-de-la-securite-technique-de-lia-les-rivaux-geopolitiques-pourraient-ils-cooperer/2026-03-25https://cesia.org/fr/publications/la-fresque-de-la-securite-de-lintelligence-artificielle-deux-heures-pour-comprendre-les-enjeux-et-les-risques-de-lia/2026-03-25https://cesia.org/fr/publications/la-securite-par-la-mesure-une-revue-systematique-de-la-litterature-sur-les-methodes-devaluation-en-securite-de-lia/2026-03-25https://cesia.org/fr/publications/lancement-du-centre-pour-la-securite-de-lia-cesia/2026-03-25https://cesia.org/fr/publications/tenir-le-cap-sur-lai-act/2026-03-25https://cesia.org/fr/publications/publication-du-code-de-bonnes-pratiques/2026-03-25https://cesia.org/fr/publications/les-dialogues-de-turing-notre-serie-de-table-rondes-sur-les-enjeux-de-lia/2026-03-25https://cesia.org/fr/publications/les-dialogues-de-turing-reprennent/2026-03-25https://cesia.org/fr/publications/lappel-mondial-a-etablir-des-lignes-rouges/2026-03-25https://cesia.org/fr/publications/nous-publions-le-premier-chapitre-de-notre-cours-les-capacites-des-ias/2026-03-25https://cesia.org/fr/publications/nous-publions-le-second-chapitre-de-notre-cours-les-risques-des-ias/2026-03-25https://cesia.org/fr/publications/nouveau-benchmark-article-et-billet-de-blog-de-lequipe-de-recherche-du-cesia-la-lecon-amere-de-la-detection-des-mesusages/2026-03-25https://cesia.org/fr/publications/pour-une-ia-francaise-tournee-vers-lavenir/2026-03-25https://cesia.org/fr/publications/pourquoi-un-accord-international-etablissant-des-lignes-rouges-a-lia-est-necessaire-et-realiste/2026-03-25https://cesia.org/fr/publications/recommandations-du-cesia-sur-le-projet-final-du-code-de-bonnes-pratiques-de-lue-pour-lia/2026-03-25https://cesia.org/fr/publications/masterclass-ia-medias/2026-03-25https://cesia.org/fr/publications/retour-sur-le-projet-grasp-construire-les-fondations-de-la-cartographie-de-la-securite-de-lia/2026-03-25https://cesia.org/fr/publications/symposium-sur-la-surete-de-lia-le-14-octobre-a-sorbonne-universite/2026-03-25https://cesia.org/fr/publications/il-est-urgent-de-definir-des-lignes-rouges-a-ne-pas-franchir/2026-03-25https://cesia.org/fr/publications/appel-international-pour-des-lignes-rouges-sur-lia/2026-03-25https://cesia.org/fr/publications/feuille-de-route-pour-une-initiative-internationale-en-r-d-de-lia-avancee/2026-03-25https://cesia.org/fr/publications/gouvernance-de-lia-priorites-pour-la-cooperation-internationale/2026-03-25https://cesia.org/fr/publications/le-cesia-denonce-la-decision-de-la-commission-de-simplifier-lai-act-cest-un-precedent-sur-lequel-lindustrie-pourra-sappuyer-pour-affaiblir-le-droit-europeen/2026-03-25https://cesia.org/fr/publications/le-cesia-denonce-le-decret-de-donald-trump-visant-a-neutraliser-les-lois-des-etats-sur-lia-une-fuite-en-avant-dangereuse-au-service-des-geants-de-la-tech/2026-03-25https://cesia.org/fr/publications/le-sommet-de-linaction-pour-la-securite-de-lia/2026-03-25https://cesia.org/fr/publications/plan-daction-continent-de-lia-le-cesia-appelle-la-commission-a-tenir-le-cap-dune-regulation-ambitieuse/2026-03-25https://cesia.org/fr/publications/vers-une-expertise-francaise-en-evaluation-de-lia-le-role-strategique-de-linesia/2026-03-25https://cesia.org/fr/about/2026-03-26https://cesia.org/fr/careers/2026-03-25https://cesia.org/fr/donate/2026-03-25https://cesia.org/fr/legal/2026-03-27https://cesia.org/en/2026-03-28https://cesia.org/en/our-work/2026-03-27https://cesia.org/en/publications/2026-03-25https://cesia.org/en/publications/bells-benchmark-project-to-assess-the-reliability-of-llm-supervision-systems/2026-03-25https://cesia.org/en/publications/press-release-of-cesia-on-the-france-china-joint-statement-on-ai/2026-03-25https://cesia.org/en/publications/in-which-areas-of-technical-ai-safety-could-geopolitical-rivals-cooperate/2026-03-25https://cesia.org/en/publications/take-part-in-the-ai-safety-fresk/2026-03-25https://cesia.org/en/publications/safety-by-measurement-a-systematic-literature-review-of-ai-safety-evaluation-methods/2026-03-25https://cesia.org/en/publications/launch-of-the-centre-pour-la-securite-de-lia-cesia/2026-03-25https://cesia.org/en/publications/cesia-urges-european-commission-to-stay-the-course-on-ai-act-the-uncertainty-is-technological-not-regulatory/2026-03-25https://cesia.org/en/publications/cesia-welcomes-the-eus-gpai-code-of-practice-and-urges-ai-companies-to-sign-on/2026-03-25https://cesia.org/en/publications/the-turing-dialogues-our-series-of-round-tables-on-the-challenges-of-ai/2026-03-25https://cesia.org/en/publications/the-turing-dialogues-are-back/2026-03-25https://cesia.org/en/publications/the-global-call-for-ai-red-lines-initiated-by-cesia-is-launched-at-the-un/2026-03-25https://cesia.org/en/publications/we-are-publishing-the-first-chapter-of-our-textbook-the-capabilities-of-ai/2026-03-25https://cesia.org/en/publications/we-are-publishing-the-second-chapter-of-our-textbook-risks-landscape/2026-03-25https://cesia.org/en/publications/new-benchmark-the-bitter-lesson-of-misuse-detection/2026-03-25https://cesia.org/en/publications/for-a-french-ai-geared-towards-the-future/2026-03-25https://cesia.org/en/publications/why-an-international-agreement-establishing-red-lines-for-ai-is-both-necessary-and-realistic/2026-03-25https://cesia.org/en/publications/cesias-feedback-on-the-final-draft-of-the-eu-code-of-practice-for-gpai/2026-03-25https://cesia.org/en/publications/highlights-from-the-ai-media-masterclass-cesia-and-reporters-without-borders-bring-experts-and-journalists-together-with-yoshua-bengio/2026-03-25https://cesia.org/en/publications/a-look-back-at-the-grasp-project-mapping-ai-risks/2026-03-25https://cesia.org/en/publications/ai-safety-symposium-on-october-14-at-sorbonne-university/2026-03-25https://cesia.org/en/publications/it-is-urgent-to-define-red-lines-that-must-not-be-crossed/2026-03-25https://cesia.org/en/publications/global-call-for-ai-red-lines/2026-03-25https://cesia.org/en/publications/a-blueprint-for-multinational-advanced-ai-development/2026-03-25https://cesia.org/en/publications/governing-the-frontier-priorities-for-international-cooperation-on-advanced-ai/2026-03-25https://cesia.org/en/publications/cesia-condemns-commissions-decision-to-weaken-the-ai-act-a-precedent-the-industry-will-use-to-undermine-eu-law/2026-03-25https://cesia.org/en/publications/cesia-condemns-trumps-executive-order-preempting-state-ai-laws-a-dangerous-and-reckless-move-serving-big-tech/2026-03-25https://cesia.org/en/publications/the-summit-for-the-inaction-on-ai-safety/2026-03-25https://cesia.org/en/publications/ai-continent-action-plan-cesia-calls-on-european-commission-to-stay-the-course-on-ambitious-regulation/2026-03-25https://cesia.org/en/publications/forging-frances-expertise-in-ai-evaluation-inesias-strategic-role/2026-03-25https://cesia.org/en/about/2026-03-28https://cesia.org/en/careers/2026-03-25https://cesia.org/en/donate/2026-03-25https://cesia.org/en/legal/2026-03-27