February 13, 2019

White House AI Executive Order and DoD AI Strategy

CNAS experts weigh in.

By Robert O. Work, Paul Scharre, Gregory C. Allen, Kara Frederick, Michael Horowitz and Elsa B. Kania

This week, the U.S. government released two major new strategic documents on artificial intelligence. The White House published an executive order by President Trump launching the American AI Initiative, and the Department of Defense released a summary of the DoD AI Strategy. Experts from the Center for a New American Security (CNAS) weigh in on these developments.

White House Executive Order on AI

Gregory C. Allen, Adjunct Senior Fellow: “This is an important act by the White House in that it formalizes AI as a national priority in the upcoming FY2020 budget request and directs federal agencies to submit plans for increasing the use of AI in executing their missions. Effective follow up and follow through will be critical to sustain the momentum generated by this executive order.”

Kara Frederick, Associate Fellow: “Monday’s executive order is slim on the details, but that’s okay. Its utility is as a signaling document that 1) identifies the engines of tech power projection where the United States should focus investment (workforce development, standard setting, etc.) and 2) nests the discussion within the great power contest where it belongs.”

Dr. Michael C. Horowitz, Adjunct Senior Fellow: “AI is a key future enabler of economic and military power. The Trump administration’s new executive order on AI makes clear that competing with China and others to sustain U.S. AI leadership is a priority. How new U.S. investments in AI and interagency coordination will actually occur is less clear. A key challenge will be balancing the desire by the U.S. government to set AI priorities with the need to foster America’s comparative advantage – a more free-market approach to innovation than America’s technological competitors.”

Elsa B. Kania, Adjunct Senior Fellow: “American leadership in AI will never be assured or uncontested. The launch of this initiative is thus critical at a time when these technologies are emerging as a focus of strategic competition. The emphasis on education and workforce development in this executive order should be welcomed, since these will be vital determinants of the United States’ future national competitiveness. It is also encouraging to see recognition of the importance of promoting international collaboration in AI, since the openness and inclusivity of our innovation ecosystem has been, and should remain, core aspects of our competitive advantage.”

Brendan McCord, Adjunct Senior Fellow: “It is critical to think about what it means to articulate and instill a vision compelling enough to address this more comprehensive work. In the past, the world saw through the Apollo program what a successful project that combines the efforts of industry, academia, and government looked like. With all forces concentrated on a single object, and through vigorous development in areas of advanced technology, the United States unlocked new exploration, achieved technology breakthroughs, and inspired generations of researchers. In the modern era, however, there are analogs in AI that are very troubling – active national projects of similar scale appear to be designed to control or oppress entire populations and create tech-enabled forms of authoritarian government. What would be the purpose-driven, democratic version of the AI megaproject?”

Paul Scharre, Senior Fellow: “The White House’s new executive order on artificial intelligence finally brings the national-level attention needed to ensure American competitiveness in this critical technology. The next step will be bringing industry, academia, and the government together to craft a strategy to ensure America remains a global leader in artificial intelligence. Human capital, data-privacy legislation, and smart regulation on AI will be key to fostering continued innovation. The White House will also have to follow-through with increasing federal funding for AI research. Cannibalizing other science funding is not an adequate solution and will harm American competitiveness in the long run.”

Department of Defense AI Strategy

Robert O. Work, Senior Counselor for Defense: “The U.S. is finally starting to respond to China’s integrated national plan to become the world’s AI superpower. The executive order launching the American AI Initiative, the establishment of the National Security Commission on AI, the publishing of the Department of Defense AI Strategy, and the standup of the Pentagon’s Joint AI Center are all important parts of America’s response. Today, however, they largely represent separate and distinct actions. The real payoff will come when the executive and legislative branches combine these disparate efforts into a coherent national plan of action. Let’s hope they do so soon.”

Dr. Michael C. Horowitz, Adjunct Senior Fellow: “The Department of Defense AI strategy is a welcome new initiative. Designed to promote U.S. leadership in military applications of AI while also setting norms for responsible use, the key question will be whether implementation proceeds in a coordinated fashion. Scaling military investments in AI, whether far from the battlefield or close to the fight, while also paying proper attention to safety, will require sustained investments and high-level bureaucratic attention.”

Elsa B. Kania, Adjunct Senior Fellow: “DoD’s AI Strategy recognizes the imperative of innovation and experimentation in technologies that have the potential to transform today’s paradigms of military power. In particular, the emphasis on creating a common foundation for, and developing a world-class workforce in, AI will be critical in efforts to actualize this agenda. The call for American leadership to lead in AI safety and military ethics also reflects a recognition that our values must be at the heart of our approach.”

Paul Scharre, Senior Fellow: “It is encouraging to see military ethics and AI safety as a major pillar of DoD’s new AI Strategy. With this new AI Strategy, DoD is clearly demonstrating their intent to build ‘human-centered AI’ that is reliable, robust, resilient, and secure. To execute this strategy, DoD will need to support it with adequate funding in the forthcoming budget that demonstrates DoD is actually making AI a priority. Without sufficient funding, this strategy is unexecutable.”

Learn more.

All CNAS experts are available for interviews. To arrange one, please contact Cole Stevens at cstevens@cnas.org.

  • Commentary
    • War on the Rocks
    • July 18, 2019
    In Search of Ideas: The National Security Commission on Artificial Intelligence Wants You

    Americans don’t want to grow old wondering what happened to their country’s place in the world. U.S. global leadership has fostered international institutions, strengthened hu...

    By Robert O. Work & Eric Schmidt

  • Podcast
    • June 20, 2019
    What Could Possibly Go Wrong?

    War has been a driver of breakthrough technology for a long time. The first waves of artificial intelligence and even the internet came out of DARPA, a defense agency whose or...

    By Paul Scharre, Richard Danzig, Arati Prabhakar & Jonathan Wilson

  • Video
    • May 6, 2019
    2019 Drell Lecture: Autonomous Weapons and the Future of War

    On April 30, 2019, CNAS Technology and National Security Program Director Paul Scharre delivered the 2019 Drell Lecture on the campus of Stanford University. Scharre's remarks...

    By Paul Scharre

  • Commentary
    • Foreign Affairs
    • April 16, 2019
    Killer Apps

    The nation that leads in the development of artificial intelligence will, Russian President Vladimir Putin proclaimed in 2017, “become the ruler of the world.” That view has b...

    By Paul Scharre

View All Reports View All Articles & Multimedia