AI R&D in the FY 2023 NITRD Budget Supplement

Last week, the Networking & Information Technology R&D (NITRD) group published its annual budget supplement to the FY 2023 budget request. The supplement provides budgetary and programmatic information on the R&D investments of agencies participating under NITRD. The supplement fulfills reporting requirements under the National Artificial Intelligence Initiative Act (NAIIA) of 2020.  The FY 2023 R&D investments are categorized by 13 Program Component Areas (PCAs), including Advanced Communication Networks and Systems, Cyber Security and Privacy, Electronics for Networking and Information Technology, and Artificial Intelligence R&D. Each PCA investment is coordinated by one or more of NITRD’s Interagency Working Groups (IWGs), who promote the public-private engagements for each R&D subject area throughout the year.

For FY 2023, NITRD’s R&D budget totals $9.6B, a $1.8B increase over FY 2022 request levels. According to the supplement, the increase is primarily driven by DOD reporting AI R&D budgetary information in FY 2023.

For the remainder of this article, I’d like to focus on the AI R&D portion of the supplement.

AI R&D FY 2023 Budget Request

According to the supplement, AI R&D is defined by NITRD as the advancement of computational systems to conduct or simulate human intelligence, including “computer vision, natural language technologies, representation, learning, reasoning, recommendation, and action; novel and use-inspired application of these techniques to various domains; and examination of trustworthiness and the associated measurements, methods, and tools needed for designing, developing, and evaluating such systems.”

AI R&D strategic priorities include:

1. Making long-term investments in AI research
2. Developing effective methods for human-AI collaboration
3. Understanding and addressing the ethical, legal and societal implications and risks of AI
4. Ensuring the safety and security of AI systems
5. Development shared public datasets and platforms for AI training and testing
6. Measuring and evaluating AI technologies through metrics and standards
7. Understanding AI R&D workforce needs
8. Expanding public-private partnerships to accelerate AI advances

The AI R&D FY 2023 request totals $1.8B, a nearly 6% increase over the $1.7B requested for FY 2022. As stated above, DOD is now reporting budgetary AI R&D figures, helping give us a more complete picture of overall federal investment in the transformative technology.

Additional details:

• Agencies not listed in chart with FY 2023 AI R&D funding include NASA ($4M), Education ($2M) and NARA ($500K)
• NSF-led National AI Research Institutes total $98M for FY 2023
• DOD increase of $38M over FY 2022 driven by increases under military departments
• AI R&D funds under DARPA decreased to $44M in FY 2023 due to the ramping down of several AI programs from FY 2022 to FY 2023, including the AI Next Campaign program which ends in FY 2024
• Increase of 48% over FY 2022 at DOC driven by $14M increase at NIST to progress in solving AI-centric challenges
• Approximately $26M dedicated to foundational AI research ad new AI partnerships with Scientific User Facilities at Energy

Biden R&D Priorities:

New to the FY 2023 NITRD supplement is a list of agency investments related to OMB’s FY 2023 R&D budget priorities to fulfil the Administration’s priorities. Priorities included pandemic readiness and prevention, tackling climate change, research and innovation in critical and emerging technologies, and innovation for equity. AI programs aligned with OMB R&D FY 2023 priorities include:

Tackling Climate Change  

• Digital Twins (NOAA): Provide a virtual simulation of the ocean with computing models, data science and AI to help adapt to real world changes.
• NSF AI Institute for Research on Trustworthy AI in Weather, Climate and Coastal Oceanography (NOAA/NSF): Create trustworthy AI methods for diverse environmental science user to enhance understanding and prediction of high-impact atmospheric and ocean science phenomena, and create educational pathways to develop a diverse AI and environmental science workforce.

Research and Innovation in Critical and Emerging Technologies

• AI/ML Consortium to Advance health Equity and Researcher Diversity (AIM-AHEAD) (NIH): Establish partnerships to increase participation of underrepresented researchers and communications in the development of AI/ML models, beginning with electronic health record data.
• Bridge to AI (Bridge2AI) (NIH): Generates new flagship biomedical and behavioral datasets, develops new AI software and standards, and trains new researchers in biomedical AI.

Next in AI R&D

According to the NAIIA of 2020, NITRD, along with the National AI Office (NAIIO), must collaborate on an update to the National Artificial Intelligence Research and Development Strategic Plan: 2019 Update. The groups are also creating a report on agency progress since the 2019 AI strategic plan. Moreover, NITRD and NAIIO will continue to maintain and enhance federal resource platforms for AI researchers including the AI Research Program Repository, AI R&D Testbed Inventory, and Data and Computing resources portals.