Introduction
The federal government has introduced new policies to enhance the acquisition and utilization of artificial intelligence (AI) technologies. These policies aim to streamline processes, ensure compliance with privacy and intellectual property rights, and promote competition and innovation in AI development and deployment.
Description
Agencies are encouraged to retain the ability to reuse and share AI code and models [2], including the option to release AI code as open-source software [2]. Appropriate processes must be established to address the use of government data [2], ensuring clear ownership and intellectual property rights between the government and contractors [2]. Licensing and IP rights should be scoped based on the intended use of AI [2], with strict guidelines on the use of government data for non-contractual purposes [2]. Contractors should prepare for data rights negotiations and adhere to transparency requirements [2], as well as comply with updated procurement guidance that emphasizes the identification of high-impact AI systems.
On April 7 [1], new policies were issued to streamline the federal government’s acquisition and use of AI, replacing previous guidance to eliminate bureaucratic barriers that could hinder the adoption of AI technologies [1]. A significant change is the establishment of a single high-impact AI category [1], which includes systems that influence decisions related to civil rights [1], privacy [1] [2], and access to essential services such as education and employment [1]. For these high-impact AI systems [1], federal agencies are required to implement minimum risk management practices [1], including pre-deployment testing [1] [3], AI impact assessments focusing on privacy and civil rights [1], and ongoing monitoring for performance and potential negative effects [1].
Policies and procedures must be implemented to ensure compliance with privacy requirements when acquiring AI systems that handle personally identifiable information [2]. Agencies are directed to promote competition among AI vendors to avoid vendor lock-in [2], conducting market research and facilitating knowledge transfer upon contract termination [2]. A domestic preference for AI developed in the US is contemplated [2], though existing statutes may not adequately address this due to the classification of AI as a service [2].
Contractors are expected to increasingly utilize AI in contract performance [2], necessitating disclosure of AI use in solicitations [2]. Performance-based contracting is encouraged [2], focusing on outcome-based needs rather than restrictive requirements [2], which may enhance competition [2]. Agencies are directed to monitor and evaluate the performance of AI systems regularly [2], with contractors responsible for the outputs of their AI [2]. In line with new federal policies, agencies must document compliance with AI governance and risk management practices within 365 days and publish a public AI strategy within 180 days.
An approach to risk management emphasizes establishing minimum requirements for the responsible use of AI [2], particularly for high-impact systems that require pre-deployment testing and ongoing monitoring [3]. Contractors may need to make discretionary decisions regarding risk mitigation measures for their specific AI applications while also preparing for increased scrutiny of their systems in accordance with agency expectations. The memorandum titled “Driving Efficient Acquisition of Artificial Intelligence in Government” provides federal agencies with strategies to acquire AI technologies efficiently and responsibly [1], promoting American-made solutions and standardizing requirements across agencies [1].
Conclusion
The implementation of these new policies is expected to significantly impact the federal government’s approach to AI, fostering a more efficient, transparent [2], and competitive environment. By emphasizing risk management, privacy compliance, and innovation [3], these measures aim to ensure that AI technologies are deployed responsibly and effectively, ultimately enhancing public trust and advancing technological progress.
References
[1] https://www.mintz.com/insights-center/viewpoints/54731/2025-04-11-omb-issues-new-guidance-federal-governments-use-ai-and
[2] https://www.jdsupra.com/legalnews/kickstarting-ai-in-the-federal-2315904/
[3] https://www.jdsupra.com/legalnews/white-house-unveils-updated-ai-6679436/
