Introduction
The October 2024 Memo from the White House Office of Management and Budget (OMB) outlines significant public investments in artificial intelligence (AI) by key US government departments. It establishes new guidelines and requirements for federal agencies regarding AI procurement [4], aiming to integrate AI into government operations while ensuring safety, compliance [1] [4], and innovation.
Description
The October 2024 Memo [1] [4], issued by the White House Office of Management and Budget (OMB) on October 3, 2024, outlines significant public investments in AI by key US government departments [1], including State [1] [3] [4], Defense [1] [2] [3], Energy [1] [2] [3], and Commerce [1], as well as the Intelligence Community [1]. This memorandum establishes new guidance and requirements for federal agencies regarding the procurement of AI systems [4], which encompass any data system, software [4], application [1] [2] [4], tool [4], or utility utilizing machine learning algorithms or other forms of AI [4]. Notably, the Memo excludes AI acquired for National Security Systems or AI used incidentally by contractors [4].
As the federal government integrates AI into its operations [1], it will initiate solicitations and procurements for AI technologies [1], creating opportunities for both traditional and nontraditional government contractors [1]. Agencies are mandated to identify contracts involving “rights- or safety-impacting AI” by November 1, 2024 [4], and to modify contracts by December 1, 2024 [4], to align with the Memo’s acquisition practices [4]. In AI solicitations [4], agencies must disclose whether the AI use is rights- or safety-impacting and may include contractual terms to protect against misuse of generative AI systems [4].
The memo emphasizes the establishment of the Department of Commerce’s AI Safety Institute, which will evaluate AI models for risks and streamline the acquisition of AI for national security purposes [3]. The institute is required to conduct voluntary preliminary testing of at least two frontier AI models within 180 days of the memo’s issuance [3], following prior agreements with OpenAI and Anthropic [3]. It will serve as the primary contact for the private sector and develop guidelines for assessing AI systems related to national security and public safety [3]. Additionally, the National Institute of Standards and Technology (NIST) is tasked with developing a voluntary pre-deployment safety testing capability for advanced AI models [2], focusing on risks related to cybersecurity and personal safety [2].
Upcoming strategic documents [1], such as the National Security Strategy and National Defense Strategy [1], will provide insights into federal AI investment priorities [1]. The forthcoming National Defense Authorization Acts and Intelligence Authorization Acts will further guide contractors in resource allocation for AI-related products and capabilities [1]. By March 23, 2025 [4], agencies must adopt specific acquisition practices and contract terms for AI systems and services [4], along with submitting written notifications to OMB regarding their progress and challenges in implementing these practices [4].
The memo indicates a shift towards new procurement types [1], particularly in the Energy Department and Intelligence Community [1], which will require contractors to adapt their construction capabilities to support AI infrastructure [1]. This focus on AI is expected to stimulate demand for innovative solutions [1], attracting startups and tech firms to the government contracting space [1]. Agencies are also directed to incorporate AI mission applications in their computational facilities and facilitate the construction of AI-enabling infrastructure [2]. Contractors will face personnel challenges as the US seeks to attract global talent in AI development [1], with an emphasis on semiconductor design and production. The memo highlights the need for streamlined visa processes to facilitate access to opportunities in the American technology sector [1], underscoring the competitive landscape for recruiting top talent [1].
As AI adoption grows [1], compliance regimes will evolve [1], necessitating that contractors stay informed about changing requirements [1]. The memo addresses the need for compliance related to new internal and external threats [1], including intellectual property protection and cybersecurity measures to mitigate AI risks [1]. Agencies are required to conduct due diligence on vendors’ data supply chains and implement contractual obligations to understand AI model training processes [4]. This marks a transformative period in federal procurement related to AI technologies [1], with minimum risk management practices outlined for rights-impacting and safety-impacting AI [4], including impact assessments and ongoing monitoring [4]. Rights-impacting AI refers to AI that significantly affects individuals’ civil rights or privacy [4], while safety-impacting AI pertains to AI that impacts human life [4], the environment [4], or critical infrastructure [4].
Agencies must ensure that vendors of rights-impacting or safety-impacting AI provide necessary documentation for monitoring system performance [4], with specific testing and monitoring requirements noted for vendor accountability [4]. The memo emphasizes the need for privacy protections throughout the procurement process to safeguard civil rights and liberties [4], ensuring that AI-based biometrics avoid unreliable or unlawfully collected information and maintain sufficient accuracy in biometric systems [4]. Additionally, the Department of Defense and the Office of the Director of National Intelligence are tasked with forming a working group to make procurement recommendations [3], and covered agencies must appoint a chief AI officer and establish an AI governance board to manage technology use and risks [3]. This comprehensive approach reflects the federal government’s commitment to responsible AI procurement and deployment, while also addressing concerns regarding transparency and oversight in the deployment of AI technologies by national security agencies [3].
Conclusion
The October 2024 Memo signifies a pivotal shift in federal AI procurement, emphasizing safety, compliance [1] [4], and innovation. By setting clear guidelines and fostering collaboration between government and industry, it aims to harness AI’s potential while safeguarding civil rights and national security. This initiative is expected to drive technological advancement, attract global talent [1], and reshape the landscape of government contracting.
References
[1] https://www.jdsupra.com/legalnews/white-house-memo-on-u-s-leadership-in-3474022/
[2] https://www.millerchevalier.com/publication/white-house-memorandum-takes-ai-policy-national-security-domain
[3] https://fedscoop.com/ai-national-security-memo-avoid-squandering-lead/
[4] https://www.cov.com/news-and-insights/insights/2024/10/omb-releases-requirements-for-responsible-ai-procurement-by-federal-agencies
