This is part of an ongoing series of Covington blogs on the AI policies, executive orders, and other actions of the Trump Administration. This blog describes AI actions taken by the Trump Administration in April 2025, and prior articles in this series are available here.
White House OMB Issues AI Use & Procurement Requirements for Federal Agencies
On April 3, the White House Office of Management & Budget (“OMB”) issued two memoranda on the use and procurement of AI by federal agencies: Memorandum M-25-21 on Accelerating Federal Use of AI through Innovation, Governance, and Public Trust (“OMB AI Use Memo“) and Memorandum M-25-22 on Driving Efficient Acquisition of Artificial Intelligence in Government (“OMB AI Procurement Memo”). The two memos partially implement President Trump’s January 23 Executive Order 14179 on “Removing Barriers to American Leadership in Artificial Intelligence,” which, among other things, directs OMB to revise the Biden OMB AI Memos to align with the AI EO’s policy of “sustain[ing] and enhance[ing] America’s global AI dominance.” The OMB AI Use Memo outlines agency governance and risk management requirements for the use of AI, including AI use case inventories and generative AI policies, and establishes “minimum risk management practices” for “high-impact AI use cases.” The OMB AI Procurement Memo establishes requirements for agency AI procurement, including preferences for AI “developed and produced in the United States” and contract terms to protect government data and prevent vendor lock-in. According to the White House’s fact sheet, the OMB Memos, which rescind and replace AI use and procurement memos issued under President Biden’s Executive Order 14110, shift U.S. AI policy to a “forward-leaning, pro-innovation, and pro-competition mindset” that will make agencies “more agile, cost-effective, and efficient.”
Department of Energy Announces Federal Sites for AI Data Center Construction
On April 3, the Department of Energy (“DOE”) issued a Request for Information (“RFI”) on AI Infrastructure on federal lands owned or managed by DOE. The RFI seeks comment from “entities with experience in the development, operation, and management of AI infrastructure,” along with other stakeholders, on a range of topics, including potential data center designs, technologies, and operational models, potential power needs and timelines for data centers, and related financial or contractual considerations. As part of the RFI, DOE announced 16 potential DOE sites for “rapid [AI] data center construction,” with the goal of initiating data center construction by the end of 2025 and commencing data center operation by the end of 2027 through public-private partnerships. The comment period for the RFI closed on May 7, 2025.
President Trump Issues Executive Order on Coal-Powered AI Infrastructure
On April 8, President Trump issued Executive Order 14261, titled “Reinvigorating America’s Beautiful Clean Coal Industry,” directing the Departments of Agriculture, Energy, and the Interior to identify coal resources and reserves on Federal lands for mining by public or private actors, prioritize and expedite leases for coal mining on Federal lands, and rescind regulations that discourage investments in coal production, among other things. The Executive Order also directs the Departments of Commerce, Energy, and the Interior to identify regions with suitable coal-powered infrastructure for AI data centers, assess the potential for expanding coal-powered infrastructure to meet AI data center electricity needs, and submit a report of findings and proposals to the White House National Energy Dominance Council, Assistant to the President for Science & Technology, and Special Advisor for AI and Crypto by June 7, 2025.
House CCP Committee Releases Report on DeepSeek Concerns
On April 16, the House Select Committee on the Chinese Communist Party released its report on DeepSeek and its AI platform, titled DeepSeek Unmasked: Exposing the CCP’s Latest Tool for Spying, Stealing, and Subverting U.S. Export Control Restrictions. Stating that DeepSeek “represents a profound threat to our nation’s security,” the report found that DeepSeek sends U.S. data to the Chinese government and manipulates chatbot outputs to “align with the CCP’s ideological and political objectives.” The report also found that it was “highly likely” that DeepSeek used model distillation techniques to extract reasoning outputs and copy leading U.S. AI model capabilities in order to expedite development. The report further found that DeepSeek violated U.S. semiconductor export controls. The report called on the U.S. to expand export controls and improve enforcement, in addition to preparing for “strategic surprise” arising from rapid advancements in Chinese AI. Ultimately, the report may help to accelerate possible U.S. Government bans on DeepSeek along the lines of the Kansas ban discussed below.
GAO Issues Report on Generative AI Impacts
On April 22, the Government Accountability Office (“GAO”) issued a report on “Generative AI’s Environmental and Human Effects.” Noting “uncertain but substantial” energy consumption, carbon emissions, and water usage resulting from generative AI training and use, the report stated that lack of available information and unpredictable technological advancements posed challenges in measuring generative AI’s environmental effects. The report also identified five risks and challenges of generative AI that may result in “negative human effects on society, culture, and people”: lack of accountability, unsafe systems, lack of data privacy, unintentional bias, and cybersecurity concerns. The report highlighted policy options for policymakers to consider to address these risks, including maintaining the status quo, encouraging the use of AI frameworks like NIST’s AI Risk Management Framework, sharing best practices, and establishing AI standards.
White House Issues Request for Comment on 2025 National AI R&D Strategic Plan
On April 24, the National Science Foundation and OSTP issued an RFI on the “development of a 2025 National AI R&D Strategic Plan” (the “Plan”) that would rewrite the previous administration’s 2023 version of the Plan, with the goal of “maintain[ing] the United States’ dominance in AI while focusing on the Federal government’s unique role in AI R&D over the next 3 to 5 years.” According to the RFI, the Plan will focus on areas that “need prioritization by government because they serve national interests but may not provide immediate commercial returns” along with “novel mechanisms for research partnerships with industry and/or academia.” The RFI lists 12 examples of potential federal AI R&D priorities for the Plan, including “fundamental advances in AI algorithms,” “next-generation AI hardware and architectures beyond deep learning,” “AI standards, security, and reliability,” and “AI systems and education supporting American workers and improving workforce productivity.” Public input in response to the RFI is due on May 29, 2025.
Kansas Enacts Ban on Government Use of DeepSeek AI Models
On April 8, Kansas Governor Laura Kelly (D) signed HB 2313 into law. The law prohibits the use of state agency-issued devices by state employees to access AI “platforms of concern,” and requires state agencies that use AI platforms of concern to deactivate and delete their accounts. The law defines AI “platforms of concern” to include AI models owned or controlled by DeepSeek or any AI model directly or indirectly controlled by China, Cuba, Iran, North Korea, Russia, or Venezuela. With the enactment of HB 2313, Kansas joins at least 15 other states that have issued DeepSeek government use bans through legislation or executive actions, including bans in Alabama and Oklahoma issued in late-March.