President Biden issued the first National Security Memorandum (NSM) on Artificial Intelligence last week, recognizing that advances in the field of AI will have significant implications for national security and foreign policy. The memorandum builds on the administration’s policies to drive the safe, secure and trustworthy development of AI.
The White House directed the United States government to create systems that will ensure that the country leads in the global race to develop AI technology and to ensure that it is safe, secure and trustworthy, to leverage AI for national security purposes, and to advance international regulations and governance around AI technology. The NSM also seeks to ensure that AI adoption reflects democratic values and protects human rights, civil rights, civil liberties and privacy while encouraging the international community to adhere to the same values.
“While the memorandum holds broader implications for AI governance, cybersecurity-related measures are particularly noteworthy and essential to advancing AI resilience in national security applications,” RStreet cybersecurity Fellow Haiman Wong said in a statement.
The memorandum tasks the National Security Council and the Office of the Director of National Intelligence (ODNI) with reviewing national intelligence priorities to improve the identification and assessment of foreign intelligence threats targeting the U.S. AI ecosystem, Wong noted. A group of agencies including ODNI, the Department of Defense, and the Department of Justice, are responsible for identifying critical nodes in the AI supply chain that could be disrupted or compromised by foreign actors, ensuring that proactive and coordinated measures are in place to mitigate such risks.
The memorandum tasks the Department of Energy with launching a pilot project to evaluate the performance and efficiency of federated AI and data sources, in order to refine AI capabilities that could improve cyber threat detection, response, and offensive operations against potential adversaries, Wong said. The Department of Homeland Security, FBI, the National Security Agency, and the Department of Defence are tasked with publishing unclassified guidance on known AI cybersecurity vulnerabilities, threats, and best practices for avoiding, detecting, and mitigating these risks during AI model training and deployment, as well.
“Our competitors want to upend U.S. AI leadership and have employed economic and technological espionage in efforts to steal U.S. technology. This NSM makes collection on our competitors’ operations against our AI sector a top-tier intelligence priority, and directs relevant U.S. Government entities to provide AI developers with the timely cybersecurity and counterintelligence information necessary to keep their inventions secure,” the White House said in a statement.
These guidelines are an important step to making sure that AI is leveraged in safe, thoughtful ways for both industry and national security, Jeffrey Zampieron, distinguished software engineer at defense technology firm Raft, said in a statement. “Fundamentally, this is quality control. We want to ensure that AI behaves in a manner that is safe, and efficacious for the application of interest. Guidelines provide creators with structured consistent ways to evaluate their work and provides consumers with confidence that the AI will work as intended,” Zamperion said.
The risks of unregulated AI technologies could be severe, he said.
“Risks lead to hazards and hazards lead to harms. The primary risk is that we give AI control of some critical behavior and it acts in a way that causes harm: Physical, Property, Financial. It’s very application specific. What’s the risk of using AI to tell jokes? Not much. What’s the risk of using AI to fire ordinance? Quite high,” he said.