Leonardo shows Michelangelo, an AI missile shield for Europe (ukdefencejournal.org.uk)

Italian defence firm Leonardo unveiled "Michelangelo," a proposed Europe-wide air and missile defence architecture that fuses real-time data from radars, satellites and infrared sensors and uses AI-assisted decision-making to shorten warning and response times. CEO Roberto Cingolani framed it as a multi-domain connective layer — a distributed "kill web" replacing siloed land/air/sea doctrines — that links heterogeneous national platforms (drones, ships, tanks, satellites) into an interoperable mesh compatible with NATO standards. Technically the concept stresses extremely low-latency sensor fusion (Leonardo cites data flows in the hundreds of terabytes per second), automated sorting/forecasting to compensate for human-bandwidth limits (e.g., only seconds to react to hypersonic threats at ~20–25 Mach), and bespoke, Europe-supervised AI rather than cloud providers’ services. The company points to recent intercept exercises and its industrial portfolio (heavy-lift drones, next-gen tanks, GCAP fighter collaboration, satellite programs) as enablers. The significance for AI/ML practitioners is twofold: first, Michelangelo demands real-time, scalable ML pipelines for fusion, tracking, attribution and predictive guidance under strict latency, reliability and ethics constraints; second, it raises engineering and governance challenges around distributed inference, secure sovereign models, protocol-driven interoperability and human–AI teaming in high-stakes, time-critical decisions. Politically and logistically, Michelangelo remains an architecture rather than a deployed system — promising rapid integration and EU complementarity (Sky Shield) but dependent on sustained government funding, cross‑border data/sharing agreements and consensus on ethical guardrails.