Introduction
On October 24, 2024 [2], President Biden issued a National Security Memorandum (NSM) to enhance US leadership in artificial intelligence (AI) while ensuring its safety [2], security [1] [2] [3] [4] [5], and trustworthiness [1] [2] [4] [5]. This memorandum [2] [3] [5], mandated by the October 2023 AI Executive Order [2], outlines a comprehensive national security strategy for frontier AI models, emphasizing their implications for national security and foreign policy [5].
Description
On October 24, 2024 [2], President Biden issued a National Security Memorandum (NSM) aimed at advancing US leadership in artificial intelligence (AI) while ensuring its safety [2], security [1] [2] [3] [4] [5], and trustworthiness [1] [2] [4] [5]. This memorandum [2] [3] [5], a requirement of the October 2023 AI Executive Order [2], outlines a comprehensive national security strategy and policy regarding frontier AI models [2], defined as cutting-edge general-purpose AI systems [2]. The NSM emphasizes the significant implications of AI for national security and foreign policy [4], drawing inspiration from historical national security documents and positioning itself as a transformative strategy for technology [2].
The NSM establishes three primary policy objectives for the US national security community:
-
Energy and Infrastructure: It underscores the urgent need to develop energy and data center infrastructure to support AI advancements [2], recognizing that much regulatory authority lies outside the executive branch [2].
-
Counterintelligence: The memorandum extends counterintelligence efforts to key players in the US AI industry to safeguard against espionage and allocate resources to counter adversaries’ theft and disruption of AI technologies.
-
Governance Frameworks: It calls for the establishment of robust governance frameworks, including the designation of a Chief AI Officer and the creation of an AI National Security Coordination Group [2]. The NSM requires agencies to implement mechanisms for risk assessment [5], accountability [2] [3] [4] [5], and transparency [1] [3] [4] [5], addressing privacy invasions [4], bias [3] [4] [5], discrimination [5], and other human rights concerns [5]. The integration of Civil Liberties and Privacy Officers in AI assessments is emphasized to ensure alignment with democratic norms such as transparency and civil liberties.
The NSM prioritizes AI safety and security initiatives [2], asserting that ensuring the security and trustworthiness of AI will facilitate its faster adoption [2]. The US AI Safety Institute (AISI) is identified as the main contact for private sector AI companies regarding testing and evaluation [2], facilitating collaboration with national security agencies to ensure the safe and reliable performance of AI systems [4]. Additionally, the NSM promotes collaboration with non-traditional vendors and streamlines procurement practices to enhance the effective use of AI systems [5].
To support the development of advanced AI systems [4], the NSM highlights the importance of secure and diverse semiconductor supply chains [4], following the CHIPS Act [4]. It reinforces the National AI Research Resource to enable diverse research contributions and directs an economic assessment of the US private sector AI ecosystem [4].
The NSM introduces a “Framework to Advance AI Governance and Risk Management in National Security,” which aims to guide “high impact” AI activities that could pose significant risks to national security and democratic values [3]. It establishes minimum risk management practices for high-impact AI activities [3], including data quality assessment and bias mitigation [3], while advocating for an interagency review process for high-impact operations [3]. A waiver process is included for Chief AI Officers to bypass certain risk management practices under specific conditions [2], mandating risk management [1] [2] [3] [4] [5], accountability [2] [3] [4] [5], and transparency measures to address privacy [4], bias [3] [4] [5], and human rights concerns [4] [5].
Internationally [1] [3] [4] [5], the NSM builds on recent progress in AI governance [4], including the development of an International Code of Conduct on AI and a Political Declaration on the Military Use of AI [4] [5], which has garnered support from numerous nations [5]. It instructs the Department of State to develop strategies for advancing international AI governance norms and calls for collaboration with allies to create a responsible governance framework that adheres to international law while safeguarding human rights and fundamental freedoms. The NSM also advocates for legally binding treaties on military AI governance, recognizing the importance of international humanitarian and human rights law [3].
While the NSM makes strides toward a cohesive governance framework for military AI [3], concerns remain about the robustness of protections [3], particularly within the Intelligence Community [3]. The executive branch retains the ability to bypass policy guidance in favor of national security needs [3], and while whistleblower protections are mentioned [3], they do not substitute for comprehensive oversight [3]. The policy emphasizes the need for national security agencies to track and assess AI systems for risks and biases [3], update internal guidance [3], and ensure accountability.
Conclusion
The NSM represents a significant step toward integrating AI into national security frameworks, emphasizing safety [1] [4] [5], security [1] [2] [3] [4] [5], and international collaboration. However, it highlights the need for stronger safeguards and oversight to ensure accountability and protect civil rights. The memorandum’s implementation may be influenced by the upcoming US presidential election [2], potentially affecting its direction and effectiveness.
References
[1] https://www.ischool.berkeley.edu/news/2024/white-house-issues-new-directive-ai-and-national-security
[2] https://www.csis.org/analysis/biden-administrations-national-security-memorandum-ai-explained
[3] https://www.justsecurity.org/104242/memorandum-ai-national-security/
[4] https://www.whitehouse.gov/briefing-room/statements-releases/2024/10/24/fact-sheet-biden-harris-administration-outlines-coordinated-approach-to-harness-power-of-ai-for-u-s-national-security/
[5] https://www.presidency.ucsb.edu/documents/fact-sheet-biden-harris-administration-outlines-coordinated-approach-harness-power-ai-for
