This article is more than 1 year old

NSA hopes AI Security Center will help US outsmart, outwit, and outlast adversaries

Agency boss warns enemies trying to nick AI advances and 'corrupt our application of it'

The US National Security Agency (NSA) is trying to head off trouble early by opening an office to oversee the development and integration of AI capabilities into defense and national security systems.

According to the Department of Defense (DoD), this AI Security Center will become the focal point for developing best practices, evaluation and risk frameworks for the adoption of AI in security-critical systems, and will consolidate the agency's various artificial intelligence activities.

Discussing the move during a meeting hosted by the National Press Club in Washington, NSA director General Paul Nakasone warned it is critical for the US to maintain leadership in AI as the technology continues to develop.

"AI will be increasingly consequential for national security in diplomatic, technological, and economic matters for our country and our allies and partners," he said.

Nakasone claimed the US might currently lead in this field, but its continued leadership cannot be taken for granted. "Our adversaries, who have for decades used theft and exploitation of our intellectual property to advance their interests will seek to co-opt our advances in AI and corrupt our application of it," he said, in what appears to be a thinly veiled reference to China.

According to the Associated Press, the founding of the AI Security Center comes after an NSA study that identified the securing of AI models from theft and sabotage as a major national security challenge.

Back in August, the DoD also announced the creation of Task Force Lima, a team set up especially to investigate how generative AI models could be put to use by the military, with the stated aim of protecting national security.

This task force was created by Deputy Secretary of Defense Kathleen Hicks to analyze different tools such as large language models to explore how to integrate them into the military's systems.

A spokesperson told us the AI Security Center will not have oversight of DoD AI efforts such as Task Force Lima.

However, Nakasone said, the AI Security Center will work closely with US industry, national labs, and academia across the intelligence community as well as the DoD and "select foreign partners."

He added that the "unique talent and expertise" at the NSA made the agency well suited to support the government's effort to ensure the US maintains its competitive edge in AI.

Meanwhile, in the UK, an industry executive recently warned a Parliamentary committee that the British government's procurement procedures were not capable of dealing with the rapid pace of AI development, and projects often take so long that the tech would be outdated before entering service.

Andrew Kinniburgh, director-general for defense within manufacturers' association Make UK, told the AI in Weapon Systems Committee that procurements "regularly take years," by which time "AI has moved on immeasurably." ®

More about

TIP US OFF

Send us news


Other stories you might like