This paper presents the Cluster SequoIA’s roadmap, addressing trustworthy AI for security in digital systems, with a focus on cybersecurity, defense, and environmental applications. The work is structured around three pillars: core AI (secure, explainable, and hybrid models; lifelong learning; formal verification), AI for cybersecurity and defense (vulnerability assessment, adversarial robustness, and dynamic security policies), and AI for environment and ocean (modelling complex systems, uncertainty quantification, and heterogeneous data integration). Transversal challenges, such as continuum computing, MLOps, ethical/legal frameworks, and user-centric design, emphasize the need for interdisciplinary collaboration to ensure AI’s responsible and secure deployment in critical domains.