Army to Solicit Industry Feedback on Layered Defense Framework for AI

The U.S. Army is planning to post two requests for information in the summer to solicit industry input on the development of a layered defense framework for testing artificial intelligence, a project called #DefendAI.

At the Defense Scoop DefenseTalks conference, Young Bang, the U.S. Army’s principal deputy assistant secretary for acquisition, a 2024 Wash100 winner and a speaker at the Potomac Officers Club’s upcoming 2024 Army Summit, said the first RFI will seek to identify processes and tools that could enable the framework implementation while the second one will focus on gathering industry feedback about the framework to make necessary adjustments, Breaking Defense reported.

The RFIs will follow the launch of the 2024 Scalable AI competition facilitated by the Army outreach office xTech. The competition, focused on automated AI risk management and robust testing and evaluation of AI, aims to inform the #DefenseAI effort.

During the event, Bang and his chief aide for software Jennifer Swanson presented an overview of the framework showing an inverted pyramid that categorizes AI component areas as high, medium or low risk.

The framework requires tailored security mitigations for AI testing in high-risk systems such as weapons and tactical networks and relaxed mitigation measures for back office applications and other low-risk tools.