Computing Research Policy Blog

NITRD Releases Supplement to President’s FY23 Annual Budget, Giving the Community First Opportunity to see Details on IT and AI R&D Expenditures for FY22

The Federal Government invested $8.76 billion in Information Technology R&D in Fiscal Year 2022 (FY22) – an increase of $410 million, or 4.9 percent over the previous year – according to an Administration report issued on Tuesday. The report is a supplement to the President’s FY23 Budget Request issued by the Networking and Information Technology Research and Development (NITRD) Program, the coordinating office for the 26 member agencies of the Federal government that are engaged in IT R&D. This supplement, which is mandated by law, “provides a technical summary of NITRD members’ budget investments and activities planned and coordinated through the NITRD Program from FY 2021 through FY 2023.” For the community, the usefulness of the document is seeing how much the Federal research agencies spent in FY22 (ie: the previous fiscal year) and in which research areas; we did not have that information until now.

The report covers all 26 member agencies of NITRD, including NSF, NIH, and DARPA. That is then further broken down into “Program Component Areas,” or PCAs, which focus on key research areas.

Particularly noteworthy is the PCA covering Artificial Intelligence (AI). The report states that across all Federal research agencies, the government spent $2.58 billion on AI R&D in FY22. That represents an increase of $130 million, or +5.3 percent, over the previous fiscal year. Looking at specific research agencies, NSF ($654 million), NIH ($551 million), DARPA ($457 million), and the Dept of Defense ($391 million) were the major funders of AI R&D in FY22, according to the report, with DOE Office of Science ($130 million) close behind.

While overall the numbers for AI R&D show an increase, specific agencies did take a step back in funding levels in FY22. For example, NSF’s AI spending dropped by $28 million over the previous year and DOE SC’s spending dropped by $10 million. It’s not clear if this is a reporting issue – that research characterized as AI-related changed from year to year – or if it represented an actual reduction in investment in AI research year-over-year. When we find out we’ll update this post.

NITRD is, “the Nation’s primary source of federally funded research and development in advanced information technologies (IT) in computing, networking, and software,” and has coordinated Federal R&D investments in advanced digital technologies for over 30 years. NITRD has 26 member agencies, more than 80 participating agencies, and Federal investments in the NITRD Program, “have increased from $7.8 billion requested in FY 2022 to $9.6 billion requested in FY 2023” according to the Biden Administration. NITRD also released a Cybersecurity R&D Strategic Plan Implementation Roadmap on Tuesday.

It’s worth noting that this is just a report to Congress showing in detail the President’s FY23 Budget Request, which was released in March, for IT and AI R&D programs that are coordinated through NITRD. Congress still must appropriate funds for Fiscal Year 2023, a matter that the legislature is dragging its feet on. But this report is useful as a status check on the nation’s IT and AI research investment, which appears to be healthy and mostly growing.

Post 2022 Election Analysis: Control of Congress, Outlook for 2023, Impact on Lame Duck Session, and Finalizing FY23 Budget

The 2022 Midterm Elections have come and gone, yet most of Washington is still trying to make sense of what happened. While it’s common knowledge that Republicans underperformed, they still were able to secure a majority in the House of Representatives, having won the 5 seats they needed to secure the majority. That majority is likely to expand by a few more seats, though by how much is still uncertain. Meanwhile, Senate Democrats have likely expanded their majority in the Senate, though we will have to wait for the results of the runoff election in Georgia to be certain.

What does this mean for next year? To say the Republican majority in the House is precariously small is an understatement. Having only a 4 or 5 seat majority will mean the House Republican leadership will have no room for defections on legislative votes. And given the highly independent nature of the most conservative members of the GOP House Caucus (ie: the Freedom Caucus), we can expect many votes to be nail-bitters or even see legislation pulled from consideration. There is also an open question of how House Republican leaders will maintain control and if they will need to put down constant no-confidence revolts from their caucus; as just one example, the vote for Rep. McCarthy (R-CA) to become Speaker of the House is uncertain at this time.

Even with those challenges, due to the nature of the House as a majority-rules chamber, House Republicans are in a naturally strong political position to enact their agenda. They will also have a perch to launch investigations into the Biden Administration at will, which will likely disrupt the President and various Federal Department’s operations. But how much the House Republicans stick together will determine how successful they will be in the majority.

On the Democrat side of the House, things are only slightly clearer. With Nancy Pelosi and Steny Hoyer announcing they are stepping out of their leadership positions, the House Democrats will have brand new leaders in almost all of their top roles. While this is unlikely to cause the same problems that the GOP is dealing with, there is still likely to be growing pains. Assuming the new leadership is able to keep their caucus together on key votes (historically not a strength of the Democratic Party), they could wield significant influence in a closely divided chamber. In one hypothetical scenario: Republican leadership, wanting to pass a yearly budget to avoid a government shutdown, may have to trade funding levels or policy riders to secure Democratic votes for passage because the full GOP Caucus won’t support the proposed budget language. But will Republican House leaders be willing to work across the aisle to pass key legislation? Time will tell.

Looking at the Senate, little is likely to change. With only a slightly expanded Democrat majority, Senate Republicans will still be in a strong position to influence how the chamber operates due to how the Senate works, which provides significant influence to the minority party. As well, the Democrat senators sitting on the political divide (ie: Senator Manchin (D-WV) and Senator Sinema (D-AZ)) are likely to still command much influence on key votes. We can expect the Senate to change very little and it will remain a legislative bottleneck.

In short, we are likely to see a very divided Congress for the next two years. Recent history says that does not bode well to getting much done. In one specific case, we can expect the annual budget process to be even more difficult than it has been in recent years.

Speaking of the budget, what does this mean for the rest of 2022 (ie: the lame duck session) and finalizing the Fiscal Year 2023 (FY23) budget? The typical political calculus would be that both sides would want to clear the deck this calendar year, so that the new Congress can focus on new items next year. That would mean finalizing the FY23 budget before the end of the year. However, there is a possibility that political calculus doesn’t hold; we’re already hearing rumblings that Congress may need to punt on the budget until next calendar year due to all the items on the lame duck session’s to-do list. Should that happen, it would mean we could expect FY23 to not be finalized until March, at the earliest.

Finally, what does this mean for science policy and research funding? With regard to policy issues that affect the research community, we shouldn’t see too much change. Frank Lucas (R-OK) is expected to become the chair of the House Science, Space, and Technology Committee; he has been a staunch ally of scientific research, with helping to champion the early parts of what became the Chips and Science Act. It’s likely we will see more efforts on research security and competitiveness with China on the committee’s agenda, but most things shouldn’t change too much.

With regard to annual funding, the outlook is less good. As shown with finalizing the FY23 budget, we can expect delays, and long delays at that. The dreaded year-long continuing resolution, where Federal agency accounts are funded at the previous year’s levels, could be even more likely over the next two years. But if the budget is handled at some point, it is hard to tell if we will get the Chips and Science Act funding levels, or a repeat of last year’s final budget numbers. Republicans are much more bullish on increasing defense spending, while keeping non-defense accounts flat (or cut them); we should expect that to continue and we will need to modify what we emphasize in our message to lawmakers accordingly.

Still many unknowns exist. We’ll have to see how this Congress handles the FY23 budget in the next few weeks; that will give us our first indication of what to expect. After that, the 2023 calendar year is likely to see a lot of political gridlock. And, of course, the 2024 Presidential Election is already gearing up. We’re in for a long two years.

U.S. Must Follow Through on NSF Funding or Risk Falling Behind Competitors, Says Former CRA Board Chair

In an opinion piece published in The Hill, Dan Reed, former CRA Board Chair and current National Science Board Chair, and Darío Gil, a member of the National Science Board and senior vice president and Director of IBM Research, say that Congress must follow through on fully funding NSF or risk dooming the CHIPS and Science Act.

As regular readers will recall, in July, Congress passed, and President Biden quickly signed into law, the CHIPS and Science Act. In that law, Congress authorized NSF’s budget to double in five years; within that increase, the Research & Related Activities (RRA) account and the Directorate for STEM Education (EDU) would double and triple, respectively. The law also codified the new Directorate for Technology, Innovation, and Partnership (TIP) as a part of RRA and authorized it to be a $4.1 billion account within RRA. These actions were meant to expand NSF’s mission and position the agency as the nation’s lead science agency for innovation and to maintain the country’s competitiveness in new research fields and technology.

The issue is this law only authorized these budgets numbers; they did not appropriate the funds. Put another way, the CHIPS and Science Act set Federal policy that NSF should be funded at these levels. But Congress would still need to appropriate the funds in the annual Federal budget process. Hence Reed and Gil’s opinion piece, insisting that Congress must follow through on these funding goals in order to reap the full benefits of the law.

In their piece, Reed and Gil make the case that this is not just about research funding, but also about training a STEM workforce for the future. “NSF is America’s STEM talent agency; it is poised to develop the STEM workforce at all educational levels to help meet expected shortfalls in talent, including 600,000 cybersecurity jobs, 1.1 million jobs in the bioeconomy, and 3 million additional AI workers.” They also point out several recent examples of how underfunding NSF has led to leaving quality research on the table, and how doing that negatively impacts parts of the country.

Ultimately, this is about the economic and national security health of the country. As Reed and Gil say, “an underfunded NSF holds the nation back. When the U.S. underinvests in discoveries, talent, and translation, we erode America’s ability to invent the future and risk both our economic competitiveness and national security.”

White House Announces New Blueprint for an AI Bill of Rights

The Biden Administration, acting through the Office of Science and Technology Policy (OSTP), released yesterday a set of principles aimed at creating a “Blueprint for an AI Bill of Rights.” The goal of the blueprint is to, “help guide the design, development, and deployment of artificial intelligence (AI) and other automated systems so that they protect the rights of the American public.”

From Left to right: Dr. Alondra Nelson, Deputy Assistant to the President and OSTP Deputy Director for Science and Society; U.S. Equal Employment Opportunity Commission Chair Charlotte Burrows; Secretary of Health and Human Services Xavier Becerra; Consumer Financial Protection Bureau Director Rohit Chopra; and Secretary of Education Miguel Cardona.

Within the proposal, OSTP, “identified five principles that should guide the design, use, and deployment of automated systems to protect the American public in the age of artificial intelligence.” Those principles, also called “common sense protections,” are: Safe and Effective Systems; Algorithmic Discrimination Protections; Data Privacy; Notice and Explanation; and Human Alternatives, Consideration, and Fallback. As Rohit Chopra, Director of the Consumer Financial Protection Bureau, pointed out at the rollout event, “fifteen years ago we thought these systems would take bias out; now we know that bias is baked in.”

The white paper includes a technical companion on how to move this plan from “principles to practice,” and an application framework. OSTP says these protections should be entitled to everyone in America and that this blueprint, “offers a vision for a society where protections are embedded from the beginning, where marginalized communities have a voice in the development process, and designers work hard to ensure the benefits of technology reach all people.”

OSTP’s proposal contains no enforcement policies on technology companies and doesn’t recommend any follow-on legislation. But the white paper does highlight existing federal agencies’ commitments to rule making and actions of those agencies to studying the specific impacts of these technologies.

This proposal caps a year-long effort by OSTP to collect feedback and views from the general public, technology companies, and the research community. In fact, CRA’s Computing Community Consortium submitted a formal response to OSTP’s Request for Information on updating the National Artificial Intelligence Research and Development Strategic Plan.

Given the concerns voiced over data privacy and consumer protections by members of Congress in recent years, the issue of algorithmic bias, and how to mitigate or avert it, will not go away here in official Washington. The computing community is well positioned to take part in these efforts and debates, contributing its technical expertise. In fact, Alondra Nelson, Deputy Director for Science and Society at OSTP, said at the briefing, “all of us have a role to play.”

PCAST Releases Report on Revitalizing the Nation’s Semiconductor Ecosystem, with Investments in a National Research Agenda a Major Component

On the heels of the Chips and Science Act passage into law in July, the President’s Council of Advisors on Science and Technology (PCAST) released a report outlining actions President Biden and the Department of Commerce (DOC) should take to, “maximally leveraging the historic $11 billion commitment appropriated for semiconductor R&D,” in the legislation. PCAST’s recommendation can fit into four broad categories: building a broad coalition, focusing on education and the future workforce, fostering innovation, and setting a national research agenda.

The Chips and Science Act provides funding for the Commerce Department to invest $11 billion over 5 years in semiconductor R&D (there is an additional $39 billion the law appropriates for financial assistance for domestic semiconductor manufacturing). Within that, DOC is to establish a National Semiconductor Technology Center (NSTC) and a National Advanced Packaging Manufacturing Program (NAPMP). Both of these programs figure heavily in the report’s recommendations (particularly 1, 2, 6, 7, and 10).

Of most note to the research community are recommendations 3, 4, 8, and 9 of the report:

3. The Secretary of Commerce in coordination with the Director of the National Science Foundation (NSF) should support the establishment of a national microelectronics education and training network by the end of 2023 and allocate funding on the order of $1 billion over the next 5 years to upgrade educational laboratory facilities, support curriculum development, and facilitate hiring of faculty into this field.

4. The Secretary of Commerce should ensure that NSTC-funded research…supports on the order of 2,500 scholarships and research assistantships per year across the educational spectrum.

8. The Secretary of Commerce should ensure that the NSTC founding charter allocates a significant portion of the annual funding, on the order of 30 to 50 percent, to directly fund a national research agenda. This research agenda should be broad in nature and address the following areas: materials, process, and manufacturing technologies; packaging and interconnect technologies; energy-efficient computing and domain-specific accelerators; design automation tools and methods; semiconductor and system security; and semiconductors and life sciences.

9. The NSTC should identify a set of nationwide grand challenges that are enabled through collaboration across the NSTC industrial membership and NSTC-funded research. These grand challenges should span three complementary areas that would benefit from large-scale nationwide collaboration: advanced computing into the zettascale era; significantly reducing design complexity; and proliferating semiconductors in life sciences applications.

The report, in concert with other actions that the Biden Administration has taken, demonstrate that the Biden Administration is moving quickly to implement the Chips and Science Act. But that’s only the “chips” half of the Chips and Science Act; the research policy community here in Washington is working on getting the “science” part of the legislation fulfilled. We are telling Congress to follow through on the research policy provisions of the law and fully fund the research agencies, especially NSF. They can start with the Fiscal Year 2023 budget.

FY23 Appropriations Update: A Continuing Resolution is Coming but When will it Be Passed into Law?

Despite major legislation passing into law this summer, Congress has been slow in finishing up the yearly Federal budget (ie: Fiscal Year 2023). The delay has been over a number of topics unrelated to science research, such as energy-permitting regulations and supplemental funding for things like Ukrainian military assistance. But with the fast-approaching November Midterm elections, Congress is not interested in shutting down the government just before facing voters at the polls. Hence the need for a continuing resolution (CR). Released Monday night, it would fund the government at Fiscal Year 2022 levels till December 16th. Congress has until Friday to pass it to avoid a shut down; while unlikely to happen, a government shut down can never be completely ruled out.

Once we see the results of the ballot box in November, we will have a better idea of how the budget will be finalized. If Republicans win back control of both chambers of Congress, they may to want to delay final consideration until next calendar year, when they have more control of the process. If Democrats retain control of the legislature, then they will likely want to finish the budget before the end of this calendar year. And if there is a split, it’s hard to tell exactly how matters will play out. Keep checking back for more updates.

Dr. Arati Prabhakar Confirmed as Next Presidential Science Advisor

Today the United States Senate confirmed Dr. Arati Prabhakar as the next Presidential Science Advisor and Director of the Office of Science & Technology Policy (OSTP). Nominated in June by President Biden, Dr. Prabhakar will be also serve as the President’s Chief Advisor for Science and Technology, a co-chair of the President’s Council of Advisors on Science and Technology (PCAST), and a member of the President’s Cabinet.

CRA released a statement applauding the President’s nomination of Dr. Prabhakar back in June. We look forward to working with the new Science Advisor and her team to advance the nation’s computing and information technology research.

White House Action on Open Access & Ending 12-Month Embargo on Federally Funded Science Articles

Last week the White House, acting through the Office of Science and Technology Policy, released guidance to federal research agencies to, “update their public access policies as soon as possible to make publications and research funded by taxpayers publicly accessible, without an embargo or cost.” Current policy allows journal publishers the option to embargo for 12 months (ie: put behind a paywall) any scientific paper they publish which is tax-payer funded. This new guidance from the White House will do away with that embargo period and require the articles to be publicly available immediately. All federal research agencies must update their policies for this to take effect by 2026.

Along with the announcement, OSTP explained the Biden Administration’s actions in a White House Blog post. Much of the justification for this is in a recent report OSTP released, titled “Economic Landscape of Federal Public Access Policy”.

This new guidance will likely have wide ranging impacts on the research and scientific publishing communities. If past events are any guide, there will also be an extensive variety of views from the research community on whether these actions are good or bad news. CRA is monitoring the situation and is reaching out to our membership and allies within the scientific community to gauge the impacts of these new guidelines.

FY23 Appropriations Update: Senate Appropriators Provide Generous Numbers to DOE Research Accounts

Continuing our coverage of the Fiscal Year 2023 (FY23) federal budget process, we turn to the Senate Appropriations Committee’s Energy and Water bill. This bill contains the budgets for the Department of Energy’s Office of Science (DOE SC) and ARPA-E, as well as funding for the Exascale Computing R&D program, for which DOE is the lead federal agency. Similar to their CJS bill, which covers NSF, NIST, and NASA, the Senate appropriators were quite generous to the energy programs, providing larger increases than the House appropriators’ mark or the President’s March budget request.

The bill proposes an 8.3 percent increase for the Office of Science over FY22 enacted levels, bringing the agency’s budget to $8.10 billion for FY23 (an increase of $625 million). That is $100 million more than the House mark and $300 million more than the President requested. Within the Office of Science, the Advanced Scientific Computing Research (ASCR) program, which houses the majority of the computing research at DOE, would see an increase of 3.8 percent – going from $1.04 billion in FY22 to $1.08 billion for FY23.

Much like last year, that 3.8 percent increase for ASCR is deceptive and contains good news. ASCR’s research subaccounts received increases and together would increase by an average of 10 percent. However, the Exascale Computing Program, which is winding down several large construction projects, received a large 40 percent cut from its FY22 levels. But this decrease is planned, as it is what the Administration requested. So, from a research perspective, this plan is better than the overall number would suggest. This mirrors what the House appropriators did in their FY23 mark.

In the committee’s report, the Senate appropriators voiced their support for the agency’s efforts in artificial intelligence and quantum information sciences. The appropriators also spoke highly of the Office of Science’s engagement plans with HBCUs and MSIs, particularly the agency’s RENEW and FAIR programs, in order to build research capacity and further workforce development.

Finally, the Advanced Research Projects Agency – Energy, or ARPA-E, would receive $570 million for FY23, a 27 percent increase (+$120 million) in comparison to FY22. While that is slightly above the House number, it is below the President’s requested level.

FY22 FY23 PBR FY23 House FY23 Senate $ Change % Change
DOE SC Total $7.48B $7.80B $8.00B $8.10B +$625M +8.3%
ASCR $1.04B $1.07B $1.05B $1.08B +$40M +3.8%
ARPA-E $450M $700M $550M $570M +$120M +27%

Unfortunately, as mentioned with the Senate’s CJS bill, these generous numbers are not likely to be passed into law. That’s because these budgets only represent Senate Democrats’ plans and have little input or support from Senate Republicans. Because of the chamber’s 50-50 split, any legislation requires at least some minority party support in order to be passed into law. We’ll have to let the budget process play out more before we have an idea of what final FY23 numbers will look like.

In the meantime, with the beginning of the new Fiscal Year fast approaching (October 1st), a Continuing Resolution (CR) is looking to be the next step in the process. A CR would continue the funding of the Federal agencies at last year’s levels, buying Democrats and Republicans time to work on negotiations. What happens with those negotiations will depend greatly on the outcome of the Midterm Elections in November. Please check back for more updates.

FY23 Appropriations Update: Strong Budget Numbers for NSF, NIST, & NASA from Senate Appropriators; but How Likely Are They to Become Law?

In our continuing series looking at the Fiscal Year 2023 (FY23) budget, we turn to the Senate Appropriations Committee. Last week, the Democrats on the committee released their slate of appropriations bills, all at once. While this is significantly earlier than in past years, it’s also a political move: the bills only represent the Democratic priorities for the federal budget and have no input from Senate Republicans. We’ll get to the implications of that in a moment but what it means in the immediate is that the numbers for many of the research agencies that the computing cares are quite good, on the surface. In point of fact, the numbers in the Commerce, Justice, Science (CJS) bill, which contains the budgets for the National Science Foundation (NSF), the National Institute of Standards & Technology (NIST), and National Aeronautics and Space Administration (NASA), are among the most robust and generous.

NSF, in particular, is a big winner under the Senate plan. The Foundation’s topline budget would receive $10.34 billion, an increase of $1.5 billion (+17 percent) over the FY22 number ($8.84 billion). That is almost as generous as the President’s 19 percent requested increase from March, and almost double the House’s mark of +8.9 percent.

Getting into the details of the Senate’s plans, the Research and Related Activities (R&RA) account, home to NSF’s research portfolio, would receive $8.30 billion for FY23. That would be an increase of $1.10 billion increase over FY22 enacted levels, or +15.3 percent. The Directorate for Education & Human Resources account (EHR), soon to be renamed the Directorate for STEM Education (EDU) (the name gets approval from Senate Appropriators in the committee’s explanatory statement), would receive a similarly generous increase of 28.7 percent, going from $1.01 billion in FY22 to $1.30 billion, with the Senate’s FY23 mark (+$290 million).

FY22 FY23 PBR FY23 House FY23 Senate $ Change % Change
NSF Total $8.84B $10.50B $9.63B $10.34B +$1.50B +17%
R&RA $7.20B $8.43B $7.71B $8.30B +$1.10B +15.3%
EHR/EDU $1.01B $1.38B $1.25B $1.30B +$290M +28.7%

Looking at policy items in the Committee’s report, there are several topics of interest. First, the Senate appropriators, much like their House counterparts, voiced their support for the newly created TIP Directorate and provided, “up to $200 million for the Regional Innovation Engines” program, saying the committee, “believes that NSF Engines will be transformative for communities across the country.” Additionally, the Committee supports the agency’s efforts in artificial intelligence and QIS, providing funding up to the President’s request in these areas. Finally, in a nod to the need for more regional diversity in research funding, the Senate appropriators provided EPSCoR, NSF’s program to build research capacity in states that do not receive the majority of the agency’s research funding, with $250 million, an increase of $35 million over FY22 enacted levels. The committee believes, “that good ideas and high-quality research are not bound to certain geographical areas but exist across the country.”

With regard to the other research agencies in this legislation, the National Institute of Standards & Technology’s (NIST) would receive a similarly healthy increase. The Senate’s plans for the agency would include a large increase, going from $1.23 billion in FY22 to $1.70 billion in FY23, a $470 million dollar increase (or +38 percent). Likewise, the Science and Technical Research and Services (STRS) account, where the majority of the agency’s research is housed, would see a push-up of +15 percent, going from $850 million in FY22 to $975 million for FY23, an increase of $125 million. While the Senate’s topline mark is well above the President’s request, the number for STRS is exactly what the Administration requested for the account.

FY22 FY23 PBR FY23 House FY23 Senate $ Change % Change
NIST Total $1.23B $1.48B $1.47B $1.70B +$470M +38%
STRS $850M $975M $953M $975M +$125M +15%

Finally, looking at NASA’s budget, the space agency would receive $26.0 billion, an increase of $1.9 billion from FY22 (+8.3 percent). The agency’s Science account would receive $8.00 billion, an increase of $431 million from FY22 (+5.3 percent). These numbers are more generous than the Administration’s request or the House appropriators’ mark.

FY22 FY23 PBR FY23 House FY23 Senate $ Change % Change
NASA Total $24.00B $26.00B $25.50B $26.00B +$1.90B +8.3%
Science $7.60B $7.99B $7.91B $8.00B +$431M +5.3%

Now, back to the politics question: do these numbers have a chance of becoming law? Unfortunately, it’s not likely. It goes back to the 50-50 split of the Senate: for any piece of legislation to move in the chamber, it requires some Republican support. Seeing as these bills and their budget numbers were put together without Republican support, it makes it very difficult for these bills to be passed into law. In all likelihood, we will have a repeat of the Fiscal Year 2022 budget, namely long months of negotiations and then final budget numbers that are lower than any of the individual plans.

In the near term, we can definitely expect a Continuing Resolution to get us through the first few months of the approaching Fiscal Year (which begins on October 1st). After that, we must wait and see what happens in the Midterm Elections. The results of those elections will likely dictate when the final appropriation bills, or a single omnibus bill, or some combination, will get a final vote. We’ll be tracking events closely, so please keep checking back for updates.

Please use the Category and Archive Filters below, to find older posts. Or you may also use the search bar.