Understanding US Federal AI Policy: recommitting to a transatlantic coalition on AI
When it comes to artificial intelligence (AI), the United States is one of the most influential countries. Due to the culmination of cutting-edge research, a strong culture of entrepreneurship and venture capital to match, and successful tech companies that have attracted leading minds from all over the world, the US is home to many of the largest companies who develop AI technologies. What is more, the American Government has also taken a stance on AI that seeks to leverage this leadership, which is clearly communicated in the AI strategy, “Artificial Intelligence for the American People.” In order to best understand the American AI policy landscape, it is imperative to understand what is happening at the Federal level in the US. Domestic AI policy in the United States has significant implications for international AI policy, and the European foreign policy community must understand not only how American tech companies influence the global AI landscape, but also how Federal policy in the US may affect AI policy all over the world.
The American goal of technological supremacy is not new, nor is it exclusive to AI. Artificial intelligence is, however, a technology that has caused geopolitical tensions in recent years and has led to the development of a national strategy, which reiterates the aspiration towards American leadership in an aggressive way. This notion of American leadership in developing and producing AI is propagated in every statement relating to AI – be it a presidential Executive Order, a federal agency document, or even rhetoric used by leading advisors on artificial intelligence. The Trump Administration’s goals, statements, and targets for AI all put one philosophy at the forefront: America First. Though Trump may be louder or more aggressive with his rhetoric, the longevity of America’s strength in the global tech realm makes the case that it is unlikely that anything should drastically change even under a Biden Administration. The US will likely still present itself as a hegemon in emerging technologies. America is pushing a whole-of-government approach to achieve and maintain American leadership in AI technologies. The persistence and priority of this goal is what is important for European policymakers to understand in order to find synergies between the European AI initiative and the American AI hegemon.
AI Policy under Obama
Before the Trump Administration’s AI agenda, the Obama Administration opened the conversation on AI policy at the Federal level with a set of documents. In October 2016, the National Science and Technology Council (NSTC) (Committee on Technology) released their “Preparing for the Future of Artificial Intelligence.” This was a 40+ page document outlining important policy areas that should be addressed by the Federal Government going forward. These areas included Applications of AI for Public Good; Research and Workforce; Fairness, Safety and Governance; and Global Considerations and Security. The NSTC released a second document entitled, “The National Artificial Intelligence Research and Development Strategic Plan” to set their research and development (R&D) intentions for AI. This R&D plan laid out seven concrete strategies, and two concrete recommendations that were to advise the Administration’s efforts on AI. The advisors to the Obama White House included issues like understanding and addressing the ethical, legal, and societal implications of AI; ensuring the safety and security of AI systems; developing effective methods for human-AI collaboration; and more.
Although these two documents were not an official American national AI strategy, they may have had a strong global influence over other countries’ national AI strategies. Some Chinese policy experts speculate that the Chinese government misinterpreted these documents as an official American AI strategy and scrambled to put together an equally strong and ambitious Chinese AI strategy in turn.So, even though these two documents were not intended to serve as an official national AI strategy, they were perceived as such from abroad, and may have been a catalyst for countries all over the world to release their official AI strategies. However, when Trump entered the White House, the new Administration super-charged these documents with America First rhetoric and laid the foundation for an American advanced technology nationalism that may easily outlive him.
Trump’s American AI Ambition
After two years in Office (and significantly later than many European countries), the Trump Administration replaced the documents created under the Obama Administration with newer, more ambitious, and more national AI aspirations. In February of 2019, Trump signed the Executive Order on “Maintaining American Leadership in AI.” This Executive Order lays the groundwork for the coordinated Federal Government strategy, the so-called “American AI Initiative.” The NSTC is still a key, cabinet-level body for Trump’s AI endeavors, but the Trump Administration created a new committee within the NSTC to spearhead the Federal AI activities: the NSTC Select Committee on Artificial Intelligence (“Select Committee,” created 2018). Together with the Select Committee, Trump’s Administration has also identified the key Federal agencies who will play a larger role in shaping, developing, and implementing American AI. These key agencies include: the National Institute of Standards and Technology (NIST), the National Science Foundation (NSF), the Defense Advanced Research Projects Agency (DARPA), the Intelligence Advanced Research Projects Activity (IARPA), the Department of Commerce, and the Department of Defense. Together, these agencies will coordinate and push the US Federal efforts on AI. This “whole-of-government" approach is pushed through two main strategies in tandem: R&D and technical standards.
AI R&D strategy
The American AI Initiative implements a “whole-of-government" strategy in order to effectively collaborate and engage with the private sector, academia, the public, and like-minded international partners. While this goal may seem global and cooperative in scope, the language written into the updated National AI R&D Strategic Plan makes it clear that the goal is to enable the United States to drive discovery and remain a world leader in AI. This R&D Strategy was commissioned first by the Obama Administration in 2016, but the Trump Administration had it updated in 2019. In doing so, it added an eighth strategy and included more nationalistic rhetoric to the existing seven. The new strategy calls for expanding “public-private partnerships to accelerate advances in AI.” This specific strategy emphasizes another key aspect of the American AI Initiative: its relationship to the private sector. This focus of the updated Federal R&D strategy stresses that Federal agencies need to strengthen public-private partnerships in AI R&D by leveraging investments and expertise, and “In doing so, the U.S. Government will capitalize on a uniquely American innovation ecosystem.” The strategy pursued today by the Federal Government is not necessarily dissimilar to that pursued under Obama. The strong focus on private sector and increased R&D is a continuation of what Obama started. The major divergence, however, is the weight given to the exceptionalist goal of putting “America First” and protecting this leadership. Looking at the rhetoric coming from opinion pieces and speeches from the US’s Chief Technology Officer (appointed by President Trump), Michael Kratsios, it is clear that enhancing public-private partnerships is planned to be done in parallel with lowering regulatory hurdles. In an op-ed in January of 2020, Kratsios promotes a “light-touch regulatory approach” and writes that “The White House is directing federal agencies to avoid preemptive, burdensome or duplicative rules that would needlessly hamper AI innovation and growth.” Advances in AI technologies have been largely driven by the American private sector. This is perhaps one reason that the government is trying to leverage this speed and innovation through public-private partnerships rather than playing catch-up. These partnerships will be crucial to the future of the American AI landscape, and the AI capabilities in both the private and public sectors.
AI standards strategy
An AI policy from the Federal Government that pushes for American Leadership can be seen just by looking at the title of the August 2019 plan from NIST, U.S. Leadership in AI: A Plan for Federal Engagement in Developing Technical Standards and Related Tools. This plan seeks to advise how the US government can play an active and purpose-driven role in AI standards development. Like the updated R&D strategy, the plan on AI standards calls for expanding public-private partnerships and strategically increasing participation in the development of technical standards. It is important to remember that technical standards are very often international in scope, so this strategy looks at how the Federal government can play a meaningful role to “champion” US standards priorities and activities around the world. Like the R&D strategy, the standards strategy keeps the same light-touch regulatory goals, especially so that American companies can continue to have an international market share. The plan emphasizes how Federal engagement with AI standards is a must, since standards developed without the right degree of Federal engagement may in fact exclude or disadvantage both US companies and US government agencies. Language used in this plan makes it clear, that while it is important for the US to engage with international parties, this should only be done if and when this engagement advances the priorities and AI standards that push US economic and national-security needs. The strategy makes clear that while cooperation may be necessary (and possible), it must always put America first.
With each individual piece of the American AI Initiative, from R&D ambitions to how the Federal government should engage at standards developing organizations, the America First rhetoric becomes clearer and more aggressive. The groundwork being laid by the government is positioning the United States for a sustained future of innovation in AI, regardless of the political tides. Although Trump is not the first President to aim for tech hegemony – and likely will not be the last – it is increasingly important that European policy makers understand which American Federal policies are being pushed, as this will certainly set the tone for the future of transatlantic cooperation.
This national AI strategy embodies a strong America First sentiment that is seen in many areas of the Trump Administration’s rhetoric, not just its policy on emerging technologies. Though, with a technology as influential as AI, this strong rhetoric cannot be overlooked as just talk, and must instead be understood by other countries. Trump’s American AI goals will most certainly have global implications and will likely leave a legacy that surpasses his time in Office. The true test will be how European policymakers and European foreign policy choose to engage with the US AI landscape. Strong and renewed transatlantic cooperation could be beneficial for both American and European AI. European foreign policy needs to prepare for transatlantic relations which mirror the tone set by Trump’s Administration. Cooperation is possible, but it may need to be on American terms.
Europe needs to strike a balance between complying with American needs, while still accounting for European goals and values. Leveraging the advances coming from the American private sector in a way that benefits the European tech landscape has been the status quo for years. As long as diplomatic relations don’t break down in standardization bodies and international fora, the America First rhetoric may be all talk and no action. Even under a Biden Administration come January, European foreign policy must decide how to best engage with the American AI landscape. The GDPR is a good example of how a European regulation has been able to hold American players to higher standards that reflect the values of a European market. This must be adapted and modernized to fit the current AI landscape. European policy makers, especially those focusing on foreign policy, should have a coordinated strategy or approach to deal with America as a tech hegemon of the future, where the private sectors simply may not be subject to the same regulatory red-tape, and the federal landscape will always give more weight to American economic- and national- security. If these terms are clear, cooperation can be mutually beneficial. AI is a globalized technology that relies on inputs, leading minds, and use cases from all over the globe. There are certain avenues European policy makers can follow to leverage these learnings from the American Federal AI landscape. For example, European policy makers can encourage or increase the number of transatlantic research partnerships or consortia; they can engage more in traditional diplomatic fora, or focus on strategic engagement through “Tech Diplomacy”; they can foster regular and sustained exchanges of leading experts between the two regions; and they can engage more with the American AI context to understand the benefits and advantages of the “whole-of-government" approach and learn how to best apply these positive aspects in the European-policy context in order to strengthen the European AI Ecosystem.