The establishment of an artificial intelligence security centre is underway within the auspices of the National Security Agency, according to an announcement made by the soon-to-be-retiring director. This endeavour is considered mission-critical in light of the surging development and incorporation of artificial intelligence capabilities into the defensive and intelligence systems under U.S. control.
Head of the Agency, Army Gen. Paul Nakasone, detailed that the new AI centre would function as a branch of the NSA’s Cybersecurity Collaboration Centre. This division’s work primarily involves cooperative ventures with private industries and international partners, aimed at reinforcing the U.S. defense-industrial construct against threats predominantly arising from China and Russia.
Nakasone iterated during the National Press Club meet that America’s current superiority in the AI sector must not be flippantly disregarded. In particular, the escalating threat posed by Beijing warrants attention. Even as the issue of Russian or Chinese interference in the upcoming 2024 Presidential elections was broached, Nakasone confirmed no such attempts had been identified thus far.
Meanwhile, allegations suggest China has amplified its cyber operations targeting U.S. and its allied institutes, potentially embedding disruptive malware in military communications. An alert was issued by the U.S. and Japan concerning Chinese hackers setting their sights on governmental, industrial, telecommunications and other military-supporting bodies.
On the question of utilizing AI for automating threat analysis and flagging alerts, Nakasone noted the existing use of AI in intelligence and defense sectors, but stressed that decisions ultimately rely on human discretion. Despite AI’s assistance, the final call rests in human hands.
The genesis of this AI security center is a consequence of recent NSA studies that underscored the critical national security challenge posed by securing AI models from theft and sabotage, especially in the era of emerging generative AI technologies.
Nakasone expounded that the center will serve as the NSA’s apex for foreign intelligence insights, formulating best practice guides, principles, evaluation methodology, and risk frameworks concerning AI security and encouraging the secure development and utilization of AI across national security systems and the defense industrial base.
Collaborative work with U.S. industry, national labs, academic institutions, the Department of Defense and international partners is on the anvil. With Nakasone’s impending retirement, Air Force Lt. Gen. Timothy Haugh is set to assume dual charge of the NSA and U.S. Cyber Command, entrusted with the leadership of both U.S. cyber defense and offense alongside signals intelligence through telecommunications surveillance.
