Federal AI and IT Research and Development Spending Analysis

Federal investment in Research and Development (R&D) to support transformative advances in Artificial Intelligence (AI), networking, supercomputing, and big data continues to grow within traditional R&D-focused agencies and to expand across government. Figures included in the FY25 budget were released on November 27, when the National Science and Technology Council (NSTC) within the Office of Science and Technology Policy (OSTP) completed its Congressionally-required supplement to the President’s FY25 budget. NSTC is tasked with ensuring that science and technology programs and policy decisions across government align with the President’s objectives. For this budget supplement, NSTC relied on two of its subcommittees: the Machine Learning and Artificial Intelligence Committee (MLAI), with analysis done by the National Artificial Intelligence Initiative Office (NAIIO) established by the National Artificial Intelligence Initiative Act of 2020, and the Network and Information Technology Research and Development (NITRD) program, which was founded under the terms of the High Performance Computing Act of 1991.

Now in its 33rd year, the NITRD Program continues to provide the R&D foundation that promotes enduring U.S. technological leadership to meet the Nation’s needs for advanced information technologies…[and] will help ensure that IT continues to be a major driving force for the Nation’s economy and security, and the health and prosperity of its citizens.

NITRD Press release
November 27

RESOURCE TOTAL AND TOP AGENCIES

The NSTC reports resources and activities across government for programs that represent significant information technology R&D and foundational work in artificial intelligence. As seen in Chart I, annual and supplemental funding for fiscal years 2021- 2025 show an increase of $2.8 billion over the five-year period, growth of 6% per year. The amounts include all reported funding for federal agency IT and AI R&D programs, including supplemental funds, except for classified programs.

*Classified program funding not included in NITRD Report. Chart I: Total Federal IT R&D and AI Funding (Unclassified), FY21-FY25 Budget Request. Source: NITRD/NAIIO Supplements to the President’s FY23, FY24 and FY25 Budgets.

As we saw last year, the agencies with the biggest IT and AI R&D footprints are the National Science Foundation (NSF), the National Institutes of Health (NIH), the Department of Defense (DOD), the Defense Advanced Research Projects Agency (DARPA) (counted separately from DOD), and the Department of Energy (DOE).

NSF’s $2.05 billion FY25 request includes the largest increase for any agency (+$285 million) over FY24 enacted levels. The increase is primarily in AI-focused research (from $400 million to $494 million), with $12-27 million increases planned for seven other categories including Advanced Communications Networks and Systems and Large-scale Data Management and Analysis (see Chart III  for the entire list of investment categories).

At $3.05 billion, NIH accounts for 27% of all IT and AI R&D funding, with a focus on Large-scale Data Management ($1.4 billion) and High-Capability Computing Infrastructure and Applications (HCIA) ($421 million).

Department of Defense (excluding DARPA) funding is planned for $2.035 billion in FY25. Its resources are balanced across functions, with the majority of its funds dedicated to Advanced Communication Networks and Systems ($408 million) and Intelligent Robotics and Autonomous Systems ($298 million). DARPA, reported separately, has a $1.41 billion FY25 budget. Over half of DARPA resources are dedicated to R&D for Electronics for Networking.

Another big player is the Department of Energy, which accounts for $1.54 billion in the FY25 budget; HCIA and Enabling R&D for HCIA account for $1.03 billion of DOE’s total.

Other agencies expecting to invest over $100 million in AI and IT R&D work include NIST ($265 million), the Department of Agriculture ($216 million), NASA ($121 million), and NOAA ($106 million). Fifteen additional agencies account for $445 million in total FY25 planned spending, as shown in Chart II.

Chart II: AI and IT R&D Resources by Agency, FY24 and FY25 Budget. Source: NITRD/NAIIO Supplement FY25, FBIQ.

In addition to reporting funding by agency, NSTC organizes the IT and AI R&D spending in 12 functional categories, called Program Component Areas (PCAs), listed in Chart III. Five PCAs account for two-thirds of the FY25 budget: Large-scale Data Management and Analysis (LSDMA) is 17% of the total funding; High-capability Computing Infrastructure (HCIA) 14%; and Artificial Intelligence (AI) 18%. Enabling R&D for High-capability computing systems (EHCS) and Electronics for Networking and IT (ENIT) account for 9% each.

The largest increase over the FY24 level is $436 million, a 38% increase, for the High-capability and Computing Infrastructure PCA. The second largest increase was $153 million, an 18% increase, for Electronics and Networking. Specific information about each agency’s funding for each PCA can be found in the NITRD/NAIIO Supplement. Complete definitions of each of the PCAs can be found in NITRD’s resource center.

Chart III: Total R&D and AI Funding by PCA Functional Category, FY22-FY25 Budget. *FY23 and FY24 Amounts include base and supplemental funding. Totals may not add due to rounding. Source: NITRD/NAIIO Supplement FY25, FBIQ.

AI FUNDING SUMMARY

The AI PCA is considered “core” AI funding. To get a full picture of the effect and use of AI, the Supplemental report emphasizes that core funding is augmented by “support” funding from other categories for AI-related purposes, or cross-cutting funding, which NITRD defines as “AI Crosscut: those [activities] with primary emphases in areas other than AI, which are reported in other PCAs.” Examples from NITRD are in the accompanying text box.

Core AI R&D examples: R&D on general methods for machine vision; machine learning; cybersecurity challenges unique to AI (e.g., ability to exploit flaws in an AI system’s goals); algorithms for computational linguistics; and neuromorphic computing architectures or chips optimized for neural net.

Cross-cutting AI investments include R&D on robots (reported in IRAS); the data analysis and management ecosystem (reported in LSDMA); broad issues of human-machine interaction (reported in CHuman) and cybersecurity search (reported in CSP); and general neuromorphic computing (reported in EHCS).

NITRD reports total AI funding for FY25 is expected to be $3.316 billion, comprised of $1.954 billion in “core” AI funding and $1.361 billion in support, or “crosscut,” AI funding. Chart IV shows Core and Cross-cutting AI funding by agency for FY25. NIH stands out, with $309 million in core AI funding, supplemented by $808 million in AI that is utilized in every PCA without exception. Thus, NIH’s total AI investment for FY25 is planned to be $1.12 billion.

Chart IV: Core AI and AI Cross-cutting Funding, FY25. Source: NITRD/NAIIO Supplement FY25 and FBIQ.

For core AI funding, the top agencies in order are NSF ($494 million), DARPA ($314 million), NIH ($309 million), DOD ($233 million), DOE ($187 million), and Agriculture ($104 million).

OUTLOOK

Given the current spending cap, the makeup of the next Congress, and the challenging negotiation for final FY25 appropriations that will go into March of 2025, non-defense agencies are likely to finish the year with less money than outlined in Biden’s FY25 Budget. However, the new Administration and 119th Congress will be avidly pursuing the competitive advantage that science and technology investments bring to a nation. Investments that boost U.S. competitiveness and can be transformative to space exploration, foundational discoveries in computing, and biomedical research spurred on by AI, will have the attention of industry, support in Congress, and the interest of the White House.