DoD AI Review Explores Scalability, Federation

U.S. Department of Defense

The Office of the Under Secretary of Defense for Research and Engineering (OUSD(R&E)) held its second annual Artificial Intelligence (AI) Defense Technical Review (DTR) on July 8 and 9 at the University of Maryland (UMD) in College Park. The event, focused on "Scalability and Federation of AI," aimed to review and set a course for intelligent decision making and strategic coordination across joint and multinational interconnected forces.

The DTR event was open to the public and attracted nearly 300 registrants from government, industry, and academia. Speakers included experts, researchers, and leaders from organizations such as Google, Microsoft, AWS, IBM, and CENTCOM.

Dr. Radha Plumb, the Pentagon's Chief Digital and Artificial Intelligence Officer, provided opening remarks on the first day of the conference, followed by a fireside chat with Mr. Maynard Holliday, Assistant Secretary of Defense for Critical Technologies, and Dr. Michael Foster, the Chief Data Officer of CENTCOM, to discuss the practical applications and challenges of federated AI. University President, Dr. Darryll Pines, also joined to emphasize academic and government collaboration. Sessions covered an overview of each of the military service and DARPA AI portfolios, mitigating challenges and finding opportunities of federated AI, federated learning, and scaling AI responsibly while balancing innovation and ethics.

The conference also included interactive breakout sessions on distributed C2 at the tactical edge, AI orchestration at operational scale, trust in federated AI, and the future of federated AI in defense.

"One of significant outcomes from this year's event included the roll out of the AI Passport concept as a new distributed Artificial Intelligence (AI) federation framework, which enables multi-party software co-development," said Dr. Kim Sablon, Principal Director for Trusted AI and Autonomy at OUSD(R&E). "AI Passport seeks to enable collaboration between diverse toolchains by sharing metadata that comply with the AI Passport Metadata Standards."

Additionally, the conference rolled out a multi agent-based C2 on-demand architecture to support joint and coalition C2 AI capabilities and highlighted the need for continuous loop monitoring and updating of AI models. The event also emphasized the importance of designing for modularity and security from the start, as well as the co-design of hardware, network, and software.

"DTR provided an opportunity for the community to participate in a discussion on the future of AI integration in defense operations and implementing a networked force powered by scalable AI solutions," said Sablon. "The discussions are of critical importance to national security, informing the development of responsible and ethical use of AI in defense operations."

/Public Release. This material from the originating organization/author(s) might be of the point-in-time nature, and edited for clarity, style and length. Mirage.News does not take institutional positions or sides, and all views, positions, and conclusions expressed herein are solely those of the author(s).View in full here.