US Public Policy and Law Alert
Artificial Intelligence (AI) continues to drive dialogue among policy makers, industry leaders, and the broader public and presents increasing congressionally authorized federal spending opportunities. For decades, the federal government has made a concerted effort to invest in the research and development (R&D) of AI. Agency budgets now total just over US$2 billion of discretionary funds allocated to funding AI projects and partnerships. As both the House and the Senate prepare to take up 12 appropriations bills before the end of the year, AI R&D funding will be under consideration. Congressional hearings on AI since ChatGPT seized the public’s attention have highlighted that Congress is looking to play a role in ensuring that the United States is at the forefront of the evolving and somewhat uncertain impacts of AI on almost every aspect of the Congressional agenda. As a result, many of the appropriations reports released by the respective Committees include numerous AI directives and funding opportunities. While only one of the actual bills, the Senate State and Foreign Operations (SFOPS) bill, includes AI within the text of the bill, nearly all of the House and Senate bill reports include clear and explicit plans for AI funding, research, and development.
Below is a brief outline of the AI directives included in each House and Senate report as well as the recommended funding which each Committee provides. Our Public Policy and Law team has been intensively engaged with lawmakers and AI experts to develop a thorough understanding of opportunities that will present themselves in these appropriations.
Both versions of the bill support the appropriated funding for the Binational Agricultural Research and Development. In the House version, the report specifically lists a US$1 million increase to expand usage of technology such as AI; however, the Senate version does not list an increase from last year in the report. Likewise, the House report includes a specific section titled AI-Empowered Specialty Crop Resilience, calling for research and optimization of AI and machine learning related to prevention of threats to specialty crops. This does not appear in the Senate report whatsoever. Yet, both reports do include a provision for the Artificial Intelligence Research Institutes, which utilizes AI to meet the future agricultural workforce needs and the Department of Agriculture’s food security mission. Similarly, both reports support using AI auditing to determine the accuracy of food labeling by the Center for Food Safety and Applied Nutrition.
COMMERCE, JUSTICE, AND SCIENCE
While the House report for the Subcommittee on Commerce, Justice, and Science has yet to be released, the Senate report already has detailed directives related to AI. Likely, the most important appropriation for AI within this report is the additional US$11 billion in new funding to continue implementation of the CHIPS and Science Act of 2022. As was the case upon passage in 2022, the goal of this funding will be to increase US competitiveness in the fields of AI, quantum computing, and microelectronics. The report also includes a dedicated AI section, which directs the National Institute for Science and Technology (NIST) to expand current AI research and development. NIST is tasked with continuing to develop AI framework for risks and trustworthiness. Likewise, the Committee states that no less than US$5 million will be used to develop AI systems and software optimization among datasets for the National Oceanic and Atmospheric Administration (NOAA). This comes as NOAA is attempting to increase data assimilation. Moreover, the Committee includes a section related to AI efforts within the National Science Foundation (NSF), and their continued funding to improve AI research and development. The recommended appropriated funds for the NSF’s research activities totals US$7.6 billion in the Senate report, of which a portion will be allocated to AI research and development grants and funding opportunities. The Senate report briefly references the implications of AI and algorithmic fairness under the Equal Employment Opportunity Commission. Part of the recommended US$455 million in appropriated funds will be allocated towards AI training and software impacts of AI on employment.
For the Defense appropriations bills, there is a significant difference in the initiatives laid out in the reports. In the Senate bill, there are multiple clear funding priorities and explicit research projects; however, in the House version, there are a few general and overarching themes around AI, but very few specifics. The Senate report has an entire section dedicated to AI, which includes recommendations for a realignment of almost US$173 million of Chief Digital and AI Officer (CDAO) resources to create a new Program Element called Alpha 1, a US$50 million increase to create an autonomy enterprise platform, and an addition US$235 million realignment for the Joint All-Domain Command and Control element. Moreover, it includes a recommendation for an additional US$141 million to implement AI at the Defense Advanced Research Projects Agency based on the National Security Commission’s recommendations. This report lays out exactly how the Senate views the changes in the Department of Defense (DOD) for AI and the specific allocations; moreover, it includes the committee’s support for ethical AI development and acceleration of programs that are currently functioning. On the other side, the House report is much less specific. While the House report mentions the importance of optimizing AI and developing clear goals for AI within the DOD, it includes very few specific recommendations. It does, however, include a US$50 million program increase for the CDAO for “modern business practices” related to the office.
ENERGY AND WATER
In the Energy and Water appropriations reports, there is not much information regarding AI application. In fact, in the House report, AI is only mentioned once, recommending a US$5 million competition for improvements in CO2 recovery through AI. The Senate report does name a few more recommendations for AI, however, including an increase in AI funding for the Dredging Operations Technical Support Program, the need for workforce development around AI, and support for Advanced Scientific Computing Research. The Senate report also has a section dedicated to AI and Machine Learning, which recommends no less than US$135 million for AI programs across the Office of Science Programs.
Due to the nature of financial services, there is a notable absence of AI-related discussion in the two appropriations reports for financial services. In the House report, AI is not mentioned whatsoever, and it is only mentioned a few times in the Senate report. One of those mentions includes potential AI opportunities to study and research potential benefits for IRS customer service programs. The only other time AI is discussed in the Senate report is, fittingly, in the AI section, which simply recognizes the potential impact of AI on society and supports the Federal Trade Commission’s actions to protect consumers from unfair or deceptive conduct through AI.
Unlike the previous appropriations bills and reports, the Homeland Security reports include more detailed applications of AI. While both chambers explicitly support development and integration of AI across the Department of Homeland Security, they take rather different approaches to the specificity of this recommendation. The House report is very detailed, recommending US$55 million for innovative border security technology like AI, US$292 million for AI target operations under the National Targeting Center, US$229 million for AI procurement related to border technology, and US$12.6 million specifically for AI in non-intrusive inspections (NII). The House report also recommends expansion of AI capabilities in ports of entry and customs facilities as well as for the Transportation Security Administration On-Person Screenings. While the House contains clear appropriations recommendations, the same is not necessarily true for the Senate. The Senate report states explicitly the desire to leverage AI to transform ports of entry security as well as a request for a briefing on prioritizing the acquisition of AI technologies for customs and border protection, but they do not give specific funding amount for AI programs. Unlike the House report, the Senate report notes a concern in the integration of AI by Immigrations and Customs Enforcement, and requests briefings to ensure AI is not abused. Finally, the Senate report recommends about US$7.5 million for AI in research and development for enhanced search and rescue operations.
INTERIOR & ENVIRONMENT
The appropriations reports for the Interior and Environment appropriations bills include minimal references to AI. Only the House report recommends AI should be used to increase accuracy in wildfire modeling by the Forest Service and Bureau of Land Management. AI is not discussed in the Senate report.
LABOR, HEALTH AND HUMAN SERVICES (HHS), EDUCATION (ED)
While the Labor, HHS, ED report for the House has yet to be released, the Senate report has a few funding directives for AI. The Senate report includes a section dedicated to AI and Machine Learning, which includes a US$135 million appropriation for the National Institute of Health’s (NIH) efforts in AI related to biomedical innovation. The Committee specifically references the NIH Strategic Plan for Data Science, which plans to increase outreach in the AI community to identify emerging biomedical topics and predict transformative breakthroughs. The Senate report also recognizes AI initiatives of the NSF and Institute of Education Sciences (IES) for AI education and research. The Committee recommends a total of US$793 million for IES funding, of which a portion will be allocated toward AI research in education.
For the Legislative Branch appropriations, there is very limited discussion of AI directives. The House report briefly references the Artificial Intelligence Working Group under the Chief Administrative Office’s investigation into increasing House efficiency through AI. Yet, the only other mention of AI within the House report is in regards to potential use of AI for captioning services during committee activity. On the Senate side, the report references the impact of AI on copyrights, urging the Copyright Office to expedite the AI initiatives. The Senate report also mentions the Government Accountability Office’s Science, Technology Assessment and Analytics (STAA) inclusion of AI for Congressional needs and relevant future STAA issues.
MILITARY CONSTRUCTION, VETERANS AFFAIRS, AND RELATED AGENCIES
The Military Construction and Veterans Affairs appropriations reports include almost no AI directives. The only mention of AI in either of the reports is a brief request for adoption of AI for Bioelectronic Medicine for Veterans within the House report. The committee urges the Department to make sure AI technologies are available for veterans.
SFOPS is the only appropriations bill which includes AI language in an actual version of the bill, not just the report. The Senate bill includes a section related to funding for democracy programs, specifically mentioning a need to address challenges for democracy associated with AI. US$159 million is made available for these specific programs, and a small portion of that will be appropriated for AI related challenges in democracy initiatives. The reports for both committees took a similar approach, as AI is not mentioned in the House report and is only briefly mentioned in the Senate report. The Senate report references AI in regard to “Technological Diplomacy” and addresses shortcomings in training for trusted AI. It also mentions the need for the North Atlantic Treaty Organization (NATO) to expand AI disinformation focus as well as increase investment in AI capabilities. The total appropriated amount for Contributions to International Organizations is US$1.44 billion, of which a small portion will be dedicated to investing in more NATO AI directives.
TRANSPORTATION AND HOUSING AND URBAN DEVELOPMENT (THUD)
In the THUD appropriations reports, there is no mention of AI, however, there are a couple of recommendations related to autonomous vehicles (AVs). In the House report, the Committee encourages the Federal Aviation Administration to develop autonomous aircraft-taxiing systems. Likewise, the House report also includes a section dedicated to “Regulatory Framework for AVs,” which directs the national Highway Traffic Safety Administration to modernize its rules for safety and timely deployment of AVs.
Current funding for the federal government will expire on 30 September 2023. Based on statements from Senate Majority Leader Chuck Schumer (D-NY) and Minority Leader Mitch McConnell (R-KY), many had expected Congress would pass a Continuing Resolution (CR) to avoid a government shutdown. Most had assumed the government would stay funded until early- to mid-December. However, passing a CR and eventually passing a year-end FY 24 Omnibus Appropriations package has proven not to be easy.
After reports that House Speaker Kevin McCarthy (R-CA) and Senate Majority Leader Chuck Schumer (D-NY) agreed to a CR from the end of September through early December, House Freedom Caucus (HFC) members stated they would not only oppose a short-term funding bill, but would vote against any legislation that does not contain their priorities: a GOP border security bill, the end of “unprecedented weaponization” of DOJ and the FBI, along with the removal of “woke” DOD policies.
As a reminder, the House has only passed one (Military Construction-VA) of 12 spending bills. Along with debate over policies and overall spending levels, there are lingering questions on procedure and timeline. Some—including members of the HFC—support the idea of rolling deadlines where a stopgap bill would be passed every couple weeks until a package is agreed upon, while others do not seem intimidated by a government shutdown.
While the appropriations bills themselves include few references to AI, the Appropriations committees may still add AI language in the various bills. With that said, it is clear from the Committee reports that AI funding and advancement is a priority for both the House and Senate. AI funding will continue to increase, and research and development opportunities will become more available. As agencies continue to prioritize AI spending, it is important to be engaged in the process from legislative ideation to agency execution. Our Public Policy and Law team is constantly monitoring both the appropriations process and the AI funding opportunities and can act as a resource for AI policy.