Artificial Intelligence (AI) and Machine Learning (ML) are increasingly leveraged for crisis management and decision support; however, their operational maturity remains constrained by fragmented data, limited adaptability, and poor interpretability. Real-world emergencies demand robust, explainable, and interoperable AI systems capable of functioning under uncertainty and degraded conditions. Recent European disasters — including wildfires (2021–2023) and floods (2021) — exposed critical shortcomings in existing systems, such as missing real-time data, non-adaptive models, and incompatible communication infrastructures that hindered coordinated response. CROSSAI addresses these challenges by developing a next-generation, AI-enabled, and fully interoperable disaster management platform that unifies technological, operational, and institutional layers across Europe. The platform integrates neurosymbolic AI engines, federated digital twins, graph and physics-informed neural networks, and explainable AI (XAI) dashboards within a secure, federated, and ontology-driven architecture aligned with GAIA-X and EU Data Space standards. This “system-of-systems” acts as a Cognitive Decision Assistant, delivering adaptive, transparent, and trustworthy recommendations to responders in real time