S18.9 · Government, Defense & Public Sector
Systems and services preparing military and government personnel, being reshaped by generative AI's adaptive scenario design.
This segment covers systems and services that prepare military and government personnel through simulation and live-virtual-constructive training. The global military simulation and training market is estimated at roughly $15.1B in 2025, with longer-range forecasts reaching $20-25B or more by the early 2030s; estimates vary considerably depending on whether the scope includes only hardware simulators or the full live-virtual-constructive training stack. The asset to watch is the content library. Simulator content libraries took years to build, and generative AI can now produce adaptive training scenarios on demand — which puts a question mark over the value of every fixed library in the segment.
Growth runs at a mid-to-high single-digit compound rate, on mandates to integrate live, virtual and constructive training and on rising adoption of VR and AR technology that substitutes cheaper simulated hours for the cost of live flight hours — a cost-out sale, which tends to be a durable one. The United States is the largest single market, driven by the Defense Department's live-virtual-constructive push; Europe follows, driven by NATO interoperability training requirements. Share concentrates among CAE, L3Harris, Leonardo DRS, Elbit and Cubic, which hold meaningful share above a long tail of VR and gaming-engine entrants.
The value chain sits downstream of the platform original-equipment manufacturers whose systems are simulated, and increasingly intersects commercial gaming-engine technology. Product type splits the economics: hardware simulators are capital-intensive and typically cost-plus, while software and VR-based training is fixed-price and carries the better margin, in the range of 15-20% or more. Revenue quality splits along the same line — simulator hardware tends to carry multi-year contracted backlog protection tied to the platform program it supports, while standalone training services are annually appropriated and competitively rebid. The entry barrier is access to platform-specific classified technical data plus facility clearance.
Neighboring markets include the simulator content built around specific platforms, commercial corporate learning, and the gaming-engine technology increasingly crossing over from entertainment into defense training. Original-equipment manufacturers are defending their proprietary simulator content; commercial gaming and VR studios are entering from entertainment technology, bringing consumer-grade production techniques — and consumer-grade cost structures — to a historically bespoke market.
Generative AI can now create adaptive training scenarios on demand rather than relying on a small number of pre-scripted exercises, and that collapses instructor and scenario-design labor along with the moat that fixed, years-in-the-making simulator content libraries used to provide. A library built over a decade is worth less once a model can generate comparable scenarios in minutes; the replacement cost of the incumbent's core asset has collapsed. A distinct category of AI-generated adaptive training content is emerging, separate from fixed-hardware simulator procurement. VR and gaming-engine adoption is already underway across the segment; AI-generated adaptive content at scale is expected within two to five years.