The report examines how we believe artificial intelligence (AI) is diffusing into defense, into the systems that armed forces, prime contractors and a new generation of startups are using to sense, decide and act faster across various domains of conflict. We explore where we believe AI is already reshaping decision advantage, warfighting and operational efficiency, and where we believe significant disruption and value creation are likely to occur next.
In the report, we analyze why the constraints on defense capability are, in our view, shifting from platforms and materials toward software, data and autonomy, and how a new generation of sensing, targeting and command systems appears to be emerging to meet that shift.
Access the Report
Report Highlights
1. Why Now
Global military spending reached $2.7 trillion in 2024, a record 9.4% increase year over year, according to the Stockholm International Peace Research Institute (SIPRI), and is projected to reach $3.6 trillion by 2030, a 33% increase from 2024 (Global X / SIPRI). Defense-tech venture investment followed: $9.5 billion invested in 2025, up 80% from $3.8 billion in 2024, per PitchBook. We examine why, in our view, this backdrop signals more than headline spending: budget attention, adoption pressure and procurement experimentation are moving in the same direction, and why our analysis focuses on the software, data and autonomy layers that increasingly determine how much mission value a platform delivers once fielded.
2. How AI Is Diffusing Into Defense
We introduce a framework for where we believe AI is creating value across defense:
- Decision Advantage, Targeting and Action, where multi-domain data assist with prioritizing assessments and recommendations;
- Warfighting, where autonomy and edge inference can be used to support mission execution in contested environments; and
- Efficiency, where AI assists with automating the administrative and maintenance work that consumes personnel time.
We use this framework to examine where the use of AIcompresses the observe-decide-act cycle without replacing operator judgment, and to assist in identifying the reliability and trust constraints that may limit how far this cycle compression can go.
3. Adoption, Procurement and the Customer Pathway
Strong technology is rarely sufficient in defense markets, and we examine why many capable defense startups fail to translate technical differentiation into fielded deployment. We analyze the shift away from the traditional program-of-record model and why, in our view, the more useful commercial question is whether a company can move from problem validation to operational deployment to funded production fast enough to matter.
4. The Defense AI Technology Stack
We map eight capability areas: autonomous systems; electronic warfare and spectrum management; ISR (intelligence, surveillance and reconnaissance); operations and domain-specific mission systems; strategic command and enterprise AI; cyber and information operations; defense manufacturing; and data and AI infrastructure. We use this stack to identify where we believe value is concentrating and where integration bottlenecks between capability areas are likely to shape which companies scale.
5. Defense AI Market Map
Our market map charts 140+ private companies across eight capability areas and 16 countries. We examine where private-market activity is concentrated, where we see whitespace and where dependency chokepoints between capability areas may determine which vendors move from pilot programs to sustained fielded adoption.