Air Force Prepares AI Training and Tools for 15 Missions

By Rebecca Stone

The Department of the Air Force will launch artificial intelligence (AI) tools for 15 missions, ranging from intelligence to nuclear operations. The Air Force and Space Force have built AI strategy and policy. They now plan to test that foundation in practice. Susan Davenport, the department’s chief data and AI officer, spoke Nov. 6 at an AI event in Washington. She said the services are increasing AI adoption while preparing troops for rapidly evolving tools. Davenport said: “What [AI] has given us is a gift. That gift is our ability to reimagine our mission to do something that we’ve never done before.”

How will the Air Force train airmen and guardians?

Davenport said the department plans to introduce a three-level approach to AI training by the end of November. New courses will teach airmen and guardians to develop and apply AI. Instruction will also address the technology’s limitations.

• The 100-level course covers general AI principles and tips for using the technology in daily work.
• The 200-level course prepares participants to build organizational workflows that use AI. Lessons go deeper on legal guardrails (rules constraining permissible uses).
• The 300-level course targets what Davenport termed hardcore developers.

The Air Force also wants to map the Defense Department’s role in AI capabilities and training, including where broader instruction ends and mission-specific instruction begins. The service will retain responsibility for general AI understanding and training on its unique missions.

Participation will be logged in troops’ permanent training records. Promotion panels and job-selection processes can consider those records as part of a service member’s overall resume. Davenport hopes eventually to measure an individual’s AI competency.

Which missions will receive AI tools?

The mission-specific rollout spans 15 areas, from intelligence to nuclear operations. Each area will receive tools designed for its specific work. A civilian or officer will oversee progress.

These tools will support flightlines, operations floors and back offices. The goal is to shorten the time required to attack a target, save commanders time and help officials make decisions faster than America’s adversaries.

The new courses will also teach personnel about AI’s limitations. Overall, the service aims to put its new AI strategy into daily operational use.

How will the department measure results?

The Air Force will assess progress through quarterly showcases. Domain principals (officials overseeing each AI mission area) will present results and seek help on persistent issues.

Two showcases have already occurred. A third is planned for this winter, in December or January.

The winter gathering will encourage collaboration among Air Force and Space Force units, research organizations and private companies, with the aim of connecting operational problems to outside technical solutions.

The first showcase revealed a gap between campaign-plan teams and research personnel. Davenport said: “Hey, hold on! I have a prototype that could meet one of your … mission outcomes.” Researchers were absent from rooms where campaign plans were discussed. That absence left mission requirements without a direct prototype proposal.

How will data handling change?

Davenport wants to move away from a single central team that cleans and tags the large volumes of data processed by the Air Force and Space Force each day.

Her proposal gives organizations greater control over their own data. Units would determine which information is reliable and share it more easily. That shift moves control toward mission organizations while supporting broader data exchange.

Davenport believes AI can transform military operations if personnel embrace the tools and apply them to missions the services have not attempted before.