- US Air Force interest in AI is expected to increase further in 2026
- Enterprise-level access to LLM and improved AI workflows are just some of the key opportunities in AI for GovCon
- Hear more actionable business intelligence about AI directly from Colonel Daniel May, Chief AI Officer of the U.S. Air Force Intelligence Division. 2026 Defense Research and Development Summit January 29th!
The U.S. Air Force's insatiable appetite for artificial intelligence is expected to grow even further in 2026. The Department of War last year Leverage AI to improve decision support, intelligence and awareness, maintenance and logistics, training and workforce capabilitiesAccording to Military.com.
Potomac Officers Club 2026 Defense Research and Development Summit The January 29th event will take a deep dive into the latest business opportunities powered by AI through engaging keynotes and engaging panel discussions. They will include senior Department of Defense officials and leading industry experts, including Col. Daniel May, Chief AI Officer, Air Force Intelligence, U.S. Air Force.
ExecutiveGov had an exclusive interview with Colonel May about the USAF's key AI business opportunities in 2026 and other AI requirements GovCon needs to know about. Get actionable business intelligence like this. 2026 Defense Research and Development Summit January 29th. Reserve your seat today!
Is the Air Force using AI?
The U.S. Air Force is experimenting with AI for a variety of applications. These include: Command and control, logistics, human-machine teaming, autonomous systems.
Let's dig into the USAF's top AI business opportunities for next year.
What is the opportunity for AI in the Air Force for GovCon?
1. Enterprise-level premium access to LLM
USAF wants to provide enterprise-level premium access to large-scale language models. May said Department of Defense organizations, including the U.S. Air Force, are currently offering free access to LLM, where users can experiment and configure their workflows.
The problem is that while this creates large bills for the USAF, it also throttles or degrades user performance, especially in accessing machine-to-machine application programming interfaces. Prime Minister May said that as a relief measure, the USAF would like to provide unrestricted paid access to at least one commercial LLM and possibly an open source LLM.
2. Make your AI models available in classified Cloud-as-a-Service
U.S. Air Force wants to expand services provided to machine learning operational pipelines that bring unclassified models into classified information networks. May said the U.S. Air Force wants to make it easy for people to create models, push a button and have them appear on a classified network for intelligence personnel to immediately access and use.
The service is already pretty good at doing this using software. May said that once a user commits the software, it is automatically pushed and deployed from unclassified networks to classified systems. Here, it is deployed on a confidential cloud computing network and offered as a service.
The USAF wants to avoid taking four months to deploy a model even though users have a better new model. May said this could lead to carriers deploying this new model to both unclassified and classified networks, and there could be two different versions of the same model.
“We don't want that to happen,” May said. “We want to be able to move, update, and fix security issues quickly, just like everyone does with software and software. [continuous integration and continuous deployment] pipeline. ”
Are you a GovCon technology expert? Then you don't want to miss Potomac Officers Club. 2026 Defense Research and Development Summit January 29th — Designed just for you. Take a closer look at our panel discussion on AI: Securing the future of defense innovation and empowering AI at mission scale. Answer your pressing questions in a Q&A session. Sign up now!
3. Improving AI workflows
The U.S. Air Force wants to help industry bring AI to workflows and optimize human-machine teamwork. It takes full advantage of both human expertise and computing technology.
The USAF is discovering how AI can benefit workflows and the steps it takes to produce something, including highlighting weaknesses. May said many organizations experience workflow weaknesses when trying to digitize processes that were done manually and on paper 20 years ago. The problem, he says, is that organizations often don't change the steps involved in their processes after going digital.
May mentioned the method IBM's Deep Blue in 1997 became the first computer to defeat a human chess champion under standard tournament rules.. IBM said Deep Blue used 32 processors to perform a series of high-speed, coordinated calculations simultaneously, examining 200 million chess positions every second.
May said that initial speculation revolved around who would win in a future human-machine chess competition: a human or a computer. He said operators found that it was neither, and instead the team perfected how to interact with the technology and understand its capabilities and limitations..
4. Applying Agentic AI to new missions
The USAF wants to apply agent AI to new missions such as foreign intelligence disclosure. This is the process by which the Department of Defense determines whether classified or controlled unclassified information should be shared with authorized representatives of foreign governments.
Prime Minister Theresa May said external information disclosure is a mission that spends a lot of time referring to policy and prior decisions to reach conclusions. He said the USAF is discovering that agent AI systems can be very useful in foreign intelligence disclosure because agents can be assigned to investigate existing policies in specific subject areas.
May also said that an AI agent could perform some collaboration, look over several documents or PowerPoint presentations, and determine whether everything is classified or categorized correctly.
“We're really starting to see AI have such a huge impact in multi-step mission areas where we need to reference a huge body of knowledge,” May said. [searching digital documents]”

