In a 2023 study, Gartner found that 65% of government CIOs said AI and ML capabilities were already in place or would be by the end of 2024 and, by the end of 2026, that number would rise to 91%. That’s a lot of expected growth over the next year or two. Efficiency, innovation and accuracy are among the reasons why.
Primer views 2025 as a turning point for AI. This year will showcase the emergence of new applications in the Department of Defense and broader adoption across other sectors. As the technology matures and trust deepens, even the most cautious industries are expected to adopt generative AI.
Here are the Top Defense AI Trends Primer has identified for 2025—both for the industry and for ourselves as we lead the charge for ever new and reliable uses for AI.
Defense AI Trend #1—Expansion of AI
Between December’s National Defense Authorization Act (NDAA) announcement and January’s Executive Order—both coming on the heels of 2023’s calls for further AI adoption by the DoD—AI is having a moment in the defense industry. It’s a critical moment, because, from here, we see the use of AI for cybersecurity, economic competitiveness and national defense exploding in the next few years.
Defense AI Trend #2—New innovations and refinements in AI
The brilliance behind AI is just beginning to scale into real-world impact. But 2025 may be the year it turns the corner. While past years have focused on foundational breakthroughs, this year will spotlight refinements that drive trust and usability in high-stakes environments. Innovations like Primer’s RAG-Verification (RAG-V) are leading that shift by enhancing AI-generated insights with source-level traceability and built-in safeguards against hallucination. These refinements will define the next era of decision-ready AI.
Defense AI Trend #3—Situational awareness powered by AI OSINT
As the Information Age accumulates more and more information, analysts and decision makers have more of a need to gain relevant insights from publicly available information (PAI) and open-source intelligence (OSINT) locked in millions of social and media posts. Using AI to aggregate and make sense of all this data is already in demand in government and industry. As more and more leverage AI for these purposes, it will become routinely used for assessing the situational landscape in real time and identifying challenges and opportunities on the path forward. If you’re not yet familiar with how AI-powered OSINT works, Primer can get you up to speed.
Defense AI Trend #4—Support for DDILs
The large language models (LLMs) that power today’s most advanced AI solutions rely on cloud computing—ideal for high-bandwidth conditions, but challenging when deployed in denied, degraded, intermittent or limited (DDIL) environments, such as in tactical military operations. Most of today’s solutions offer all-or-nothing functionality that doesn’t meet the needs of those operating in constrained or contested environments. In 2025, we see more companies following Primer’s lead of creating architectures that ensure functional AI despite limiting external conditions.
Defense AI Trend #5—User-centric semantic searching
The Boolean search methods used to scour through millions of pages of data have always presented a challenge that requires specialized skill. But what if you could just use plain English to draft a simple question, as if you were asking a colleague? We think 2025 will transform this process, making search easier and more accessible for analysts. Primer has already incorporated this capability into our products and we think the rest of the industry will follow suit in 2025. The need is too big and it’s a human-focused feature other solutions can no longer live without.
Primer is excited to see what 2025 holds for AI. We are already hard at work shaping what’s next, but if there are features or advancements you’d like to see in our products, let us know.
