AI Adoption: Are Air Force and Space Force Ready and Willing?

Air & Space Forces Magazine | Todd South

The U.S. Air Force and Space Force are accelerating artificial intelligence adoption, with Secretary of the Air Force Troy Meink announcing more than 300 AI tools for military personnel. Senior leaders want these systems to streamline administrative work, combine data and improve decision-making. Yet rapid adoption is raising concerns about oversight, security and workforce readiness.

The Department of the Air Force tracks AI applications through a central inventory called CLARA, while mission owners retain responsibility for development, authorisation and delivery. Its AI Action Team, established by Chief Master Sergeant of the Air Force David Wolfe, has examined more than 50 projects. The department also uses the Pentagon’s GenAI.mil platform for unclassified and sensitive administrative tasks. However, unapproved tools can create shadow AI, exposing data and networks while generating costs that escape central oversight. The Army’s Sage AI experience offers a warning: its annual allocation of 100 million tokens had been exhausted by mid-June after access expanded. Officials are examining usage, computing requirements and budgets, while acknowledging that training and unit-level support remain essential to effective adoption.

Comment

The central governance problem facing the U.S. Air Force is the gap between formal AI policy and everyday user behaviour. Enterprise adoption means managing tools across multiple units, networks and mission types rather than simply making software available. CLARA records relevant applications, but an inventory alone does not authorise department-wide use. The distinction matters because individual users can adopt external services faster than institutional approval processes can assess them.

The operational mechanism behind shadow AI is straightforward. Personnel seek quick solutions, enter information into accessible tools and create workflows that central IT teams may not be able to monitor. Depending on the system and information entered, this could expose sensitive data, create uncontrolled dependencies or produce duplicated spending. The Army’s Sage AI experience also demonstrates how unrestricted access can strain computing budgets. The Air Force’s own approach, including CLARA and GenAI.mil, reflects an attempt to connect experimentation with oversight, but those systems serve different functions: one tracks use cases, while the other provides access to AI tools.

Governance also depends on the people using these systems. A tool can be technically available yet remain poorly understood at squadron level, where personnel must judge whether its output fits a particular task. The AI Action Team offers a channel for identifying grassroots projects, but enterprise adoption still depends on evaluating results and managing risks. For the U.S. Air Force, the test is whether CLARA’s oversight and unit-level learning can keep pace with the spread of AI applications.

Strategic Question for Discussion

If AI use across the U.S. Air Force continues to expand faster than formal oversight, can CLARA provide meaningful visibility into the tools and workflows personnel actually use?

CLARA offers a central view of reported AI use cases, but its value depends on the completeness and accuracy of that information. The pattern suggests that enterprise governance and unit-level trust must develop together if informal AI use is to remain visible and manageable.

Share your assessment in the comments below.

To read in details click here