12 For military mission-functions, modern warfighting systems need to be able to collect data, aggregate and process it, and collaborate with humans to make decisions based on it both at an individual system level and at a command level. AI and AI-adjacent technology includes software, sensors, data architectures, and computing components that allow these digital and physical systems to operate autonomously or intelligently with a human in the loop. Chris Brose explains this throughout his book, The Kill Chain: Adjacency extends beyond warfighting functions to include the enterprise aspects of the military, because enhancing organizational and administrative efficacy can also significantly impact deterrence and warfighting outcomes.17 Fighting a war with logistical functions that are 10% faster than the enemy’s attack speed, or at a 10% lower cost to the enemy’s counter-measures can significantly change the outcome of a war, and far more preferably, deter a would-be aggressor from initiating kinetic conflict. The Fiscal Year 2025 Defense Budget allocated $1.8 billion to programs directly focused on AI, and another $1.5 billion to the Joint All-Domain Command and Control strategy, which focuses on AI-enabling technologies such as data systems, IoT, and autonomous systems. Another major AI-adjacent program is Replicator, first announced in 2023, which seeks to build “attritable”18 capabilities, meaning low-cost autonomous systems meant to field in tens of thousands of units simultaneously across multiple domains.19 Replicator received approximately $500 million in funding in 2025.20 These adjacent programs are developing and acquiring systems that will operate on computers and networks that depend on AI as the General Purpose Technology that powers many of their core operating functions, and connects them to humans. Gen-AI, specifically, can serve as a critical focal-point, processing the immense volume of data that these replicable systems intake via systems generated through Replicator, share and store on a secure network developed through JADC2, and then deliver to human operators who strategize, plan, and act upon that information. In conjunction with these funding allocations, we observe some of the largest AI companies honing their products for the defense market and forming partnerships to deliver them specifically to the military. In 2024 and 2025, several partnerships materialized between leaders in defense, technology, and AI development, including OpenAI with Anduril, Palantir with Shield AI, and Meta with Scale AI. These partnerships have several commonalities in that they compound: 1. Qualifications and experience with data aggregation and synthesis for DOD technology platforms; 2. Access to large-scale data sets and funding resources necessary to train a leading Gen-AI tool; 3. Amassment of the engineering expertise and successful RDT&E cycles necessary to develop and refine leading21 Gen-AI products, such as foundation models. While these partnerships signal a clear direction among industry players to deliver large-scale technology rollouts, the DOD’s current procurement strategy for Gen-AI products signals an approach prioritizing both large ‘prime’ awards as well as smaller contract awards, incorporating a variety of business sizes and maturity stages. Hardware will still be important, but what will more likely win future wars is information. It will be the ability to build battle networks in which every military system can connect and collaborate with all others. And the capabilities most essential to success will be artificial intelligence, machine autonomy, cyber warfare, electronic warfare, and other software-defined technologies. 16 “ Generative AI Adoption in the US Military Fighting a war with logistical functions that are 10% faster than the enemy’s attack speed, or at a 10% lower cost to the enemy’s counter- measures can significantly change the outcome of a war, and far more preferably, deter a would-be aggressor.