The goal of the new center is to advance and empower the secure adoption of AI technologies across the national security enterprise and defense industrial base, Nakasone announced, during an event hosted by The National Press Club.
“Today, the U.S. leads in this critical area, but this lead should not be taken for granted,” he said. Nakasone has served as both the commander of the U.S. Cyber Command and the director of the NSA since 2018.
Working closely with U.S. industry, national labs and academia, the new entity has been designated to be the center for developing best practices, evaluation methodology and risk framework related to AI. Nakasone said that the center would be incorporated into the NSA’s Cybersecurity Collaboration Center.
The U.S. faces threats from both China and Russia in the space of national security, Nakasone noted, as well as threats on the horizon that cannot yet be foreseen due to rapid technological advancement. He argued that AI will be increasingly consequential in the national security space.
The findings of a recently conducted NSA study on AI revealed that there is a clear need for a focus on AI security, which contributed to this action.
And while Nakasone noted that AI is already in use by U.S. intelligence and defense agencies to expand their capabilities, he expects that adversaries are moving in a similar direction while working to exploit potential vulnerabilities in U.S. and allied AI systems.
“AI security is about protecting AI systems from learning, doing and revealing the wrong thing,” he stated.
The unique expertise of the NSA qualifies the agency to support this whole-of-government approach to security, Nakasone added.
This news complements other Department of Defense (DOD) work in this space, including the January 2023 update to the DOD’s 2012 directive to govern the development of autonomous weapon systems to ensure that directive aligned with AI advances. The DOD also published the “Responsible AI Strategy and Implementation Pathway” in 2020.
