The new National Security Memorandum, which comes a year after President Joe Biden issued an executive order on regulating AI, seeks to thread the needle between using the technology to counter its military applications by adversaries and building safeguards to uphold civil rights, officials said.

"This is our nation's first ever strategy for harnessing the power and managing the risks of AI to advance our national security," National Security Advisor Jake Sullivan said in a speech at the National Defense University in Washington.

"We have to be faster in deploying AI and our national security enterprise than America's rivals are in theirs. They are in a persistent quest to leapfrog our military and intelligence capabilities."

The United States seeks to develop national security applications of AI in areas like cybersecurity and counterintelligence in an effort to curb the risk of a "strategic surprise" from its rivals, a senior Biden administration official told reporters.

"Countries like China recognize similar opportunities to modernize and revolutionize their own military and intelligence capabilities," he said.

"It's particularly imperative that we accelerate our national security community's adoption and use of cutting-edge AI capabilities to maintain our competitive edge."

Last October, Biden ordered the National Security Council and the White House chief of staff to develop the memorandum as he issued an executive order that aimed for the United States to "lead the way" in global efforts to manage the risks of AI.

The order, hailed by the White House as a landmark move, directed federal agencies to set new safety standards for AI systems and required developers to share their safety test results and other critical information with the US government.

Calls for 'transparency' 

US officials expect that rapidly evolving AI technology will unleash military and intelligence competition between global powers.

American security agencies were being directed to gain access to the "most powerful AI systems," which involves substantial efforts on procurement, a second administration official said.

"We believe that we must out-compete our adversaries," the official told reporters, adding that most of the memorandum is unclassified, while also containing a classified annex that primarily addresses adversary threats.

The memo, he said, seeks to ensure a swift adoption of the technology in a responsible way, with the government releasing a framework document alongside the plan that provides guidance on "how agencies can and cannot use AI."

In July, more than a dozen civil society groups such as the Center for Democracy & Technology (CDT) sent an open letter to Biden administration officials, including Sullivan, calling for robust safeguards to be built into the memo to protect civil liberties and ensure transparency into how intelligence agencies were deploying AI.

"It is critical that uses of AI are subject to democratic accountability notwithstanding legitimate needs for secrecy," Samir Jain, the vice president of policy at the CDT, said on Thursday.

Calling on the White House and Congress to ensure oversight, he added: "We cannot rely on national security agencies to grade their own homework."