How Congress must reform its budget process to compete against China in AI

The 30 NATO leaders were clear-eyed and direct in the Brussels Summit Communiqué, “We face multifaceted threats, systemic competition from assertive and authoritarian powers, as well as growing security challenges to our countries and our citizens from all strategic directions.” The key question is how well the United States, its allies, and partners will meet those “threats, competitions, and challenges,” especially in an era when technology is moving far faster than government decision-making.

The Biden administration’s 2022 budget proposal includes a record $112 billion for defense research and development. The increase is appropriate, but more money alone will not keep us competitive with China and other adversaries. Today’s rapid innovation and technological change renders our industrial age approach to funding obsolete. Along with spending more in crucial areas, we need significant reform in how those funds are allocated and spent.

The National Security Commission on Artificial Intelligence (NSCAI) was created to help guide Congress and the nation in one of the foundational new technologies for national security, and it documented in detail many of the problems created by the current budget process. Its Final Report provided recommendations that Congress and the executive branch should act on immediately.

To defend our country, the Department of Defense must have access to the ingenuity of the entire nation, from small innovative companies and cutting-edge academia to the traditional defense industrial base. We need all of our players on the field. But our antiquated budgeting system keeps too many from contributing.

China, of course, does not have this problem. It gains strategic advantage by mobilizing all its national resources through civil-military fusion to develop and utilize emerging technologies with artificial intelligence at the core. Of course, we should never try to imitate authoritarians, but we do need new approaches.

Over the last several years, Congress has given DOD several new authorities for acquisition and DOD has made some use of ‘work-around’ organizations like the Strategic Capabilities Office and the Defense Innovation Unit. But DOD still does not have the budgeting and funding flexibility to rapidly develop and field promising technology wherever it is developed. Although it may seem like a wonky, in-the-weeds issue, transforming how we fund the life cycle of emerging technologies, such as AI, is a strategic imperative for our nation.

Fortunately, the AI Commission has given Congress and the executive branch a roadmap that would help the Department tap into the innovative power of the American people and American small businesses more effectively than the current process.

First, the Commission recommends that Congress establish a dedicated AI fund as a near-term solution to bridge the “valley of death.” Too many times new innovations are begun but do not have an immediate sponsor within DOD, and the often-small companies cannot afford to wait through the federal budget cycle for one to materialize. This $200 million fund would be dedicated to AI and allow the undersecretary for Research and Engineering to place strategic bets on key capability-enabling technologies. The Commission does not see this fund as a panacea, nor another short-lived “slush fund,” but rather as an essential stopgap measure to accelerate AI integration now to get capabilities into the hands of the warfighters until broader budget reform can be achieved.

Second, the Commission recommends a pilot program to test a portfolio management approach for requirements and budgeting to overcome the lack of flexibility in DOD’s program-centric budget process. The pilot would test the utility of managing similar programs and systems as a group, instead of individually, to accelerate more prototyping, development, and integration of new technologies such as AI. The pilot should produce lessons learned for broader reforms to the current planning, programming, budget, and execution (PPBE) process and defense budget system.

Third, the Commission recommends establishing a single appropriation for the life cycle costs of software and digital technologies, building on an existing initiative known as the Budget Activity 8 pilot. This fund would be exempt from the traditional programming and planning process and would be used as the single source of funding for the full life cycle of capability delivery and continuous engineering.

Finally, the Commission recommends a pilot program to explore mission-focused budgeting. Mission-focused budgeting, like portfolio management, would enable more flexible movement of funds to support experimentation, early prototyping, and technology transition. In addition, it would allow moving money across services and portfolios to meet the operational needs of the joint force. Such a pilot would be tested on one operational challenge identified by the Joint Staff and one or more combatant commands.

Each of these recommendations is focused on artificial intelligence, a field which is developing at mind-numbing speed that the current industrial age budgeting process can never accommodate. These proposals provide greater flexibility to fund technologies and applications that show promise and to drop those that prove inadequate and to do it all much faster. It will be essential for all funding decisions to be fully transparent to Congress so that the appropriate oversight can be conducted. Likewise, DOD and Congress must work together to define the relevant metrics to measure effectiveness and show tangible returns on investment.

The challenge of making time-relevant decisions exists throughout the Department of Defense, but it is most acute in the newer areas of technology. If we cannot speed up decisions on AI, we have no hope of competing with China. Individually, each of these recommendations help a bit; together, they move toward a new paradigm for how the Department funds, procures, and fields emerging technology at the speed and scale required to defend the nation in a dangerous, complex world.

Thornberry served in Congress from 1995-2021. He was House Armed Services chairman from 2015–2019.