Discover the complete schedule for the AI & Cybersecurity Workshop organized by the Cluster SequoIA in Rennes, in collaboration with AI Factory France’s cybersecurity and defense vertical, on June 24–25, 2026: presentations, posters, and cross-sessions with the PEPR IA Days.
The workshop aims to provide an overview of the scientific and technological challenges at the intersection of artificial intelligence and cybersecurity, structured around three key themes:
- The robustness and security of AI systems against increasingly sophisticated attacks
- AI as a tool for cybersecurity: detection, analysis, and incident response
- New threats posed by AI to the security of digital systems.
The program, spanning June 24–25, 2026, features recent scientific research on these three major themes, offering a state-of-the-art perspective while emphasizing scientific exchange.
A Format Designed for Interaction
On June 24 , at the Couvent des Jacobins , two sessions organized in collaboration with the PEPR IA Days will kick off discussions on AI security challenges. These sessions are accessible via registration on the PEPR IA Days website, a great opportunity to explore all the work conducted under the PEPR IA initiative.
👉 Register for the joint PEPR × SequoIA Cluster session
On June 25, at the Inria Conference Center at the University of Rennes, the day will combine presentations and poster sessions, with a variety of formats designed to encourage participant interaction. To facilitate international participation, presentations will be in English.
Program Workshop AI & Security
8h30 : welcome coffee
9h - 10h40 : orals
- Léo BERTRAND | LORIA | Une architecture hybride pour agents IA de sécurité offensive : mémoire d’expérience, automatisation et exploration
- Ramzi REZKI | CNAM | Transformer-Based Multi-Class Intrusion Detection for IoT and Network Traffic Analysis
- Barbara PILASTRE | IRISA | Graphes temporels et détection d'anomalies : Application en cybersécurité
- Pierre-François GIMENEZ | Inria, IRISA | CasinoLimit: An Offensive Dataset Labeled with MITRE ATT&CK Techniques
- Victor Matrat | LORIA | When AI Confuses Syntax with Semantics: Functional Adversarial Examples Against Binary Similarity Models
10H40 - 10h50 : coffee break
10h50 - 12h10 : orals
- Valentin HEGO | Université de Rennes | Can LLM Break Obfuscation?
- Ibtissam HARROUCHE | Eurocom | Topological Privacy Vulnerabilities in Knowledge Graphs
- Romaric GAUDEL | Université de Rennes, IRISA | Privacy-Utility Tradeoffs in Decentralized Learning via Correlated Noise
- Patrick BAS | CNRS | Mixing AI and Watermarking: The Current AI Waterfall
- Alexandre DE CORNIERE | Toulouse School of Economics | AI and Cybersecurity: A Model of Vulnerability Discovery and Defense
12h15 - 13h30 : lunch break
13h - 14h40 : posters
- Yacine Tamoudi | Kereval | Hallucination Detection in Large Language Models: A Hybrid Approach for Cybersecurity Auditing
- Nicola Anselmi | Université de Rennes, IETR | AI-Driven Antenna Array Interference Mitigation: Fast Beam Scanning and Nulling for Secure Wireless Infrastructure
- Alex Pierron | Institut Polytechnique de Paris | Trustworthy And Efficient Deep Reinforcement Learning-Driven Physical-Layer For Secure 6G Networks
- Yosra Ghaouar | Développement d'un système IDS/IPS intelligent basé sur l'apprentissage par renforcement pour la cybersécurité des infrastructures IoT
- Anne Josiane Kouam | SRP, TRiBE, Inria Saclay, MLSec, TU Berlin | Les modèles de prédiction de mobilité divulguent-ils les trajectoires des utilisateurs ?
- Victor Matrat | LORIA | When AI Confuses Syntax with Semantics: Functional Adversarial Examples Against Binary Similarity Models
- Boris Zannou | Institut Supérieur de Technologie (IST) de Mamou (Guinée) | Authentification biométrique révocable : vers une résistance aux attaques par IA générative
- Billal Mokhtari | CNAM PARIS | Game-Theoretic Mitigation of Attacks in Digital Twin-Enhanced Industrial Control Systems Using Markov Decision Processes
- Cédric Prigent | LIRIS | Practical Confidential Clustering for Federated Learning using Efficient Secret-Sharing
- Antonio Cimino | Eurecom | Understanding Indirect Prompt Injections (IPIs)
- Kavé SALAMATIAN | Université Savoie Mont-Blanc | Log analysis using LLM meta-modelling
14h40 - 16h : orals
- Yufei HAN | Inria, IRISA | LLM backdoor poisoning in the finetuning process
- Antoine BOUTET | INSA Lyon, CITI | Towards the Anonymization of Masked Language Modeling Abstract
- Kavé SALAMATIAN | Université Savoie Mont-Blanc | Runtime Security for Agentic Systems
- Cédric KILLIAN | Université Jean-Monnet Saint-Etienne, Laboratoire Hubert Curien | Side-Channel Reverse Engineering of FPGA-based AI Accelerators
- Roderick ASSELLINEAU | Université de Rennes, IRISA | Cryptanalytic Extraction of Deep Neural Networks
Practical information
Seats are limited, Register here
Inria Centre at the University of Rennes, Campus de Beaulieu, 263 Avenue du Général Leclerc, 35042 Rennes
Contact: iacluster
univ-rennes [*] fr (iacluster[at]univ-rennes[dot]fr)