Xavier Lessage took part in this crisis exercise organised by the French Cybersecurity Agency (ANSSI) during the Paris AI Action Summit 2025.
Date: 11 February 2025
Event: External activities ⊕
Domaine: Manufacturing ⊕
Innovation theme: Cyber Security ⊕
About project: AIDE ⊕
On Tuesday, 11 February, Xavier Lessage had the opportunity to participate in a cyber crisis management exercise organized by ANSSI (National Agency for Information Systems Security) at La Défense (Cyber Campus). More than 100 AI and cybersecurity experts, divided into 20 teams, took part to explore concrete scenarios and anticipate the risks associated with AI, particularly in the supply chain.
The session was immersive and collaborative, highlighting the importance of resilience in the face of emerging threats and the importance of integrating security from the earliest stages of AI development.
Artificial intelligence and cybersecurity are now closely linked. At the event, major threats to AI systems, such as model poisoning, backdoors, and inference attacks, were examined. These attacks aim to disrupt the proper functioning of AI, modify its behavior in response to a given signal (such as an image containing a specific characteristic, e.g., a person wearing a hat), and ignore certain behaviors in a surveillance system. Attacks by inference, on the other hand, aim to divulge confidential information.
As part of the AIDE project and in partnership with several institutions (KULeuven/DistriNet, KULeuven/COSIC, UGENT, UCLouvain, BOSA), CETIC has been contributing since the end of 2022 by focusing on the practical usage. CETIC is also collaborating with other universities (UMons, UNamur, ULiege, ULB) and research centers (Multitel, Cenaero, Sirris, etc.) involved in similar projects.
View online : https://www.evenements.cyber.gouv.f...