Countries that can harness the current wave of innovation, mitigate its potential disruptions, and capitalize on its transformative power will gain economic and military advantages over potential rivals.

A new wave of innovation is being driven by the convergence of advances in software, physical, and biological systems.⁴ Increased automation, pervasive connectivity, and improved analytics are remaking services and manufacturing. New technologies such as genomics (comprehensive methods for studying the molecular biology of genes, cells, and physiology), additive printing (also known as three-dimensional, or 3-D, printing), and the Internet of Things are merging the physical, biological, and digital worlds.

This wave of innovation is characterized by speed, disruption, and scale. The period between technological breakthroughs is decreasing, and the pace of adoption of technologies is much faster than in the past (see figure below). After Alexander Graham Bell invented the phone, fifty years passed before half of all American homes had one; but only five years after the invention of the smartphone, half of all Americans had one.⁵ It took hundreds of millions of dollars for scientists to sequence the first human genome, in 2004; fifteen years later, machines can sequence genomes for approximately $600 each.⁶

The disruption of technology incumbents is also accelerating. In 1958, companies spent an average of sixty-one years on the S&P 500. By 2011, the average had dropped to seventeen years, and at the current rate of change, in another ten years 75 percent of companies currently on the S&P 500 will be off it. Greater agility is required to remain competitive.

The scope of change is also broader. AI and automation will cause major changes in the workforce, as a previous CFR Independent Task Force report explored.⁷ There is a great deal of uncertainty about how much job loss will result: a study from the University of Oxford predicted that 47 percent of U.S. jobs could be automated within the next decade, a McKinsey Global Institute report estimated that 23 percent of jobs could be displaced by automation by 2030, and the Organization for Economic Cooperation and Development (OECD) found that just 9 percent of jobs in the United States were at high risk. But these are gross losses, not net losses, and new jobs are likely to be created by the application of new technology. Therefore, technological change will create losers and winners. Automation tends to complement the expertise, judgment, and creativity of highly educated individuals performing technical, professional, and managerial work. It tends to displace middle-skill (often non-college-educated) workers performing routine tasks that can be coded in production, clerical, administrative, and sales roles. Even with training and educational investments, more Americans may end up working in lower-wage jobs.⁸

These overlapping characteristics make forging a successful innovation strategy more difficult and heighten the costs of failure. Nations that fall behind are likely to lack critical building blocks of economic and military power. They are less likely to have a say in how new technologies are developed and governed around the world.

The United States has led the world in innovation, research, and technology development since World War II, but that leadership is now at risk.

Before the war, federal funding for research and development was small. During the war, funding for research and development grew by a factor of twenty, large federal laboratories were founded, and federal money went to universities for research and training. As a result, scientists and engineers made contributions to the war effort, including synthetic rubber, radar, the radio proximity fuse, guided missiles, and the atomic bomb.

Fearing the loss of these efforts after the war, Vannevar Bush, the director of the Office of Scientific Research and Development and the first scientific advisor to a president, delivered a report in 1945 titled Science: The Endless Frontier that justified continued federal support for science and technology as the basis of economic and national security. According to Bush, pushing these “new frontiers of the mind” was essential “to our security as a nation, to our better health, to more jobs, to a higher standard of living, and to our cultural progress.”⁹

Bush helped formalize a “pipeline” model of radical innovation, focused on creating new-to-the-world products. The innovation process starts with federal investment in R&D, which then passes through universities that promote research and training, and ends with new products that are developed and commercialized in the private sector. Although Science: The Endless Frontier laid out a linear model starting with basic research, the reality has always been more complex, with feedback loops between the different stages.

The flow of technologies through the pipeline was often accelerated by the need to respond to political crises.¹⁰ For example, following the Soviet Union’s launch of Sputnik in 1957, the Dwight D. Eisenhower administration broadened the front of the pipeline through the National Defense Education Act, which created new graduate fellowships in science and math, provided low-interest loans to undergraduate and graduate students, supported the development of modern curricula in science and math for K–12 education, and funded training institutes for science teachers. Between 1957 and 1961, federal investment in R&D nearly doubled, and total government outlays for basic science at the National Science Foundation and other agencies almost tripled.¹¹ Congress also supported accelerants at later stages in the pipeline, establishing NASA and the Advanced Research Projects Agency (which became the Defense Advanced Research Projects Agency, or DARPA) to focus on high-risk, high-gain applied development.

Federally supported R&D had a dramatic impact on U.S. competitiveness and national security. According to a 2019 study, starting in the 2010s nearly one-third of patented U.S. inventions relied on federally funded science (see figure below).¹² Touch screens, the Global Positioning System (GPS), and internet technologies central to the smartphone are all products of Defense Department research. Department of Energy research grants played a role in the development of Tesla’s batteries and solar panels, shale gas hydraulic fracturing, light-emitting diode (LED) technology, and 4-D and 5-D seismic imaging. Grants from the NSF were important to the building of the internet, Google’s search engine, and magnetic resonance imaging (MRI) machines. Funding from the NIH drove research that supported the sequencing of the human genome, advances in prosthetics, and the cancer drug Gleevec (imatinib).¹³ Between 1988 and 2010, $3.8 billion of federal investment in genomic research generated an economic impact of $796 billion and created 310,000 jobs.¹⁴ A new wave of support for basic research could have similar economic and military benefits.

The strengths of the U.S. innovation system have been magnified by the U.S. role as a central node in a global network of research and development. Multilateral trade agreements promoted the (relatively) free flow of goods, investment, and—more recently—services and data. Federal agencies pursued a wide range of international cooperative projects on clean energy, pandemic response, food security, and other transnational challenges. U.S. companies opened R&D centers and developed research partnerships in places such as Beijing, China; Bengaluru, India; Dublin, Ireland; Haifa, Israel; Manchester, United Kingdom; and Toronto, Canada. American universities have long had large foreign student populations and partnerships with their foreign peers, and U.S. scientists were the coauthors of choice on scientific papers at higher rates than their counterparts abroad. Innovation was driven in part by the inclusive institutions and networks that connected the United States to the rest of the world.

The United States’ open, democratic system was a beacon for talented scientists and engineers from around the world. Immigrants have won half of the nation’s Fields Medals (for outstanding achievement in mathematics) and a large share of the nation’s Nobel Prizes, and they are roughly twice as likely as native-born Americans to start a new business. Sixty percent of the most highly valued technology companies today were founded by immigrants to the United States or the children of immigrants, and in 2014, one-quarter of new engineering and technology start-ups had an immigrant founder.¹⁵ eBay, Intel, and Google were all founded or cofounded by immigrants (respectively, Pierre Omidyar, Andy Grove, and Sergey Brin).

In 2017, all U.S. investment in R&D—from public and private sources—totaled $496 billion, more than in any other country. American universities dominate the global list of educational institutions advancing science, inventing new technologies, and spurring new sectors.¹⁶ Private-sector investment and R&D continue to be robust, driving new waves of invention and commercialization. Business R&D increased from $70 billion in 1980 to $300 billion in 2016, a 340 percent rise (see figure below).¹⁷ U.S. technology companies significantly outspend their competitors on R&D (see figure below). In 2017–18, U.S. firms, led by Amazon and Alphabet (Google’s parent company), invested more than $5 in R&D for every $1 spent by Chinese companies.¹⁸

Moreover, the total amount of venture capital (VC) invested in U.S.-based companies has risen from $14.5 million in 1980 to $131 billion in 2018.¹⁹ The United States attracted more than half the global investment in seed-stage funding (funding that supports proof of concept or early product development), followed by China. The United States leads the world in providing business, financial, and information services, and despite the rise of China, U.S. technology companies are the largest producer of high-technology manufacturing, with 31 percent of global share (though recently American companies have moved more of their production offshore).²⁰ In short, the U.S. innovation system retains great strengths.

Cracks in the Foundation

U.S. leadership in science and technology is at risk because of a decades-long stagnation in federal support and funding for research and development. Private-sector investment has risen, but it is not a substitute for federally funded R&D directed at national economic, strategic, and social concerns.

U.S. innovation leadership is not guaranteed. Public spending on basic science drives discoveries that would have been too big and risky for a private company to undertake. In effect, federal investment funds R&D with national economic, strategic, and social returns, while private-sector R&D is motivated by commercial returns. Moreover, public R&D creates spillovers that benefit the entire economy and incentivize greater R&D funding in the private sector. Yet despite its importance to the nation’s innovation base, federal spending on research and development as a percentage of the overall economy has declined since the mid-1980s, from 1.2 percent of GDP in 1985 to 0.66 percent in 2016 (see figure below).²¹

The end of the Cold War hastened this downward trend, as did the budget sequestration in the aftermath of the 2008 financial crisis, which mandated deep cuts to federal spending, including federal R&D. (However, research spending received a boost during the crisis, through the American Recovery and Reinvestment Act.) Federal R&D as a percentage of the budget peaked at close to 12 percent in the mid-1960s because of the Apollo space program but declined in the years after its end, to close to 3 percent in 2017.²² The main lever to increase research spending would be to reverse the long-term decline in overall domestic discretionary spending, since federal spending on civilian research as a percentage of domestic discretionary spending has been relatively steady over the last decades.

The private sector is becoming the largest funder of research, with respect to not only commercially viable applied research but basic R&D as well. In 2015, for the first time since World War II, the federal government provided less than half of all funding for basic research.²³ But the private sector invests a much smaller share of its revenues in riskier, early-stage basic and applied research than in later-stage development. In the past, corporate research laboratories such as Bell Labs and IBM Research made early-stage investments in translating basic science into products. Corporations now face short-term market and shareholder pressures to focus on incremental advances in existing technologies. They are not funding research and development that leads to new breakthroughs in science and engineering and spurs commercial growth later.

President Trump came into office with the intention of cutting the federal budget, and he has repeatedly proposed deep reductions to nondefense R&D spending, as well as to overall domestic discretionary spending. Congress has rejected these cuts, instead passing an omnibus spending bill in 2018 that provided the largest increase in R&D funding in nearly a decade, made possible by a significant increase in the cap on domestic discretionary spending and on defense spending.²⁴ The president’s 2020 budget request has again proposed sharp reductions in R&D funding, including a 13 percent cut to the NSF, a 12 percent cut to the NIH, and the elimination of Advanced Research Projects Agency-Energy, a program that funds speculative environmental technologies.²⁵

The Task Force believes the Trump administration is failing to provide the long-term leadership in R&D required to protect U.S. prosperity and security. The White House needs to recognize the urgency of the innovation challenge and reach an agreement with Congress on the budget caps for fiscal years (FY) 2020 and 2021.

Trade Disputes' Disruption of Innovation

Friends, allies, and collaborators tightly link technology ecosystems and create scale in a globalized system of innovation, and thus are a competitive advantage. Washington’s current trade policies needlessly alienate partners, raise costs for American tech firms, and impede the adoption of U.S. technology in foreign markets.

The Task Force warns that the Trump administration’s indiscriminate use of tariffs against China, as well as partners and allies, will harm U.S. innovative capabilities. The White House has rightfully targeted Beijing’s market-manipulating policies and theft of intellectual property as central issues in the bilateral relationship. But extensive tariffs on Chinese information and communications technology (ICT) products will increase component costs for U.S. companies, leaving them less money for their own U.S. ICT investment, and will thus lower productivity and slow the growth of output by perhaps $163 billion over the next ten years.²⁶ The tariffs also disproportionately target imports of intermediate inputs—products American companies purchase to combine with other materials to place in final products and then sell globally. As a result, U.S. goods are more expensive and thus less competitive against those of other producers. The trade wars also impede the adoption and deployment of American technologies in foreign markets, as Canada, the European Union, Mexico, Russia, and Turkey have all retaliated against American exports.²⁷

Moreover, Trump’s declaration that trade disputes with friends and partners are based on national security threats and his willingness to enter into intensive trade battles with friends and adversaries alike have undermined alliances and thus weakened efforts to change China’s behavior. Asian and European partners are also concerned about China’s technology threat and the erosion of their industrial base. The European Commission’s March 2019 review of the EU’s relations with China, for example, criticizes Beijing for “preserv[ing] its domestic markets for its champions, shielding them from competition through selective market opening, licensing and other investment restrictions; [and] heavy subsidies to both state-owned and private sector companies.”²⁸ U.S. tariffs on products from friends complicate, if not make impossible, any effort to build a broad coalition and simultaneously make it easier for Beijing to play trading partners against each other.

The Trump administration erred in withdrawing from the Trans-Pacific Partnership (now known as the Comprehensive and Progressive Agreement for Trans-Pacific Partnership, or CPTPP). Joining the CPTPP, a trade deal covering twelve countries and close to 40 percent of the global economy, would have strengthened U.S. leadership in Asia and increased leverage on Beijing. China would have faced pressure to conform to CPTPP trade rules in order to attract investment, and leaving the agreement weakened U.S. credibility with its Asian allies.²⁹

The United States has also gradually been decentered in global scientific and technology networks. This is due partly to the rise of Brazil, India, and other hubs of scientific discovery. But China is also competing with the United States for the roles of funder and partner, promoting scientific collaboration as part of the Belt and Road Initiative (BRI), its effort to connect to the Indian Ocean, Persian Gulf, and Europe. The Chinese Academy of Sciences, for example, has invested almost $268 million as part of the BRI and opened nine research and training centers in Africa, Central Asia, South America, and South and Southeast Asia.³⁰

The Narrowing Talent Pipeline

A central strength of the U.S. innovation environment has been a steady pipeline of domestic STEM talent and the country’s ability to attract the best and brightest students, engineers, and scientists from around the world. A lack of strong education initiatives at home and new barriers to talented foreign students’ and workers’ coming to and remaining in the United States will have long-term negative economic and national security consequences.

The United States is also seeing a decline in its ability to attract highly educated immigrants, and the number of new international students enrolling at American institutions fell by 6.6 percent during the 2017–2018 academic year, after a 3.3 percent decline the year before.³¹ These decreases have been driven by increased competition from other countries for talent, as well as by U.S. gun violence, public safety fears, and concern about restrictive immigration policies. Recent policy decisions, such as two 2017 executive orders banning travel to the United States for citizens from seven Muslim-majority countries, have complicated scientific exchange. These travel restrictions have disrupted researchers’ plans, and a number of technology conferences, such as a meeting of the Internet Engineering Task Force, have been rescheduled to venues outside the United States to allow foreign participation.³²

Actions by the Trump administration to limit H-1B visas have hampered tech firms that rely on top global talent to staff their operations. The denial rate for applicants trying to extend their visas grew from 4 percent in 2016 to 12 percent in 2018 to 18 percent in the first quarter of 2019.³³ The administration has also proposed ending the work authorizations for H-4 visa holders (the spouses of H-1B visa holders), making it yet more difficult to retain talent. In addition, in June 2017, the Department of Homeland Security (DHS) proposed ending the International Entrepreneur Rule, which provides temporary residency to foreign entrepreneurs starting a business in the United States. Other countries, such as Australia and Canada, are using these developments to lure talent. In a 2019 survey of four hundred U.S. hiring professionals from big and small companies, 63 percent said they were increasing their presence in Canada, either by sending more workers there or by hiring foreign nationals, because of U.S. immigration policies.³⁴

In May 2019, the White House announced a new merit-based immigration plan, which would replace the family-based system of green card issuance with a Build America visa favoring workers with “extraordinary talent,” “professional and specialized vocations,” and “exceptional academic track records.” The changes could result in up to 5,340 more immigrants with a master’s degree or higher getting visas.³⁵ Even if the plan were to gain support in Congress, which currently looks unlikely, the Task Force does not believe the Build America visa effectively addresses the global talent race since it does not allow immigrants highly skilled in tech to bring their families with them. Very few skilled workers will come to the United States without their families when they can choose similar opportunities in Australia, Canada, or the United Kingdom and keep their families together.

Concerns about whether the United States itself is producing enough scientists, engineers, and technologists date back to Sputnik and the competition with the Soviet Union, and have reemerged almost every decade since. Today, an increasing percentage of college graduates major in STEM fields, but significant shortfalls remain in government and industry in specific sectors such as cybersecurity, mechanical engineering, systems engineering, and aerospace engineering.³⁶ U.S. universities are currently projected to produce fewer than 30 percent of the required number of graduates to fill the 1.4 million computer specialist job openings. Moreover, a smaller number of these students are American citizens, as the proportion of foreign students studying STEM subjects in the United States has doubled in the last thirty years. One estimate is that given current trends, international students will make up half of all STEM doctorates by 2020.³⁷

It is important to note that the United States is not fully utilizing American talent, either. Minorities and women remain underrepresented in STEM fields. Only 2.2 percent of Latinos, 2.7 percent of African Americans, and 3.3 percent of American Indians and Alaska Natives hold a university degree in STEM fields.³⁸ Women constitute 47 percent of the overall workforce but only 28 percent of the science and engineering workforce, and women in tech jobs leave the field at a rate 45 percent higher than men.³⁹ More inclusive environments not only would address talent shortages and inequality of opportunity, but also are essential to economic and national security since they are demonstrated to be more innovative.⁴⁰

The Defense Department and the intelligence community will fall behind potential adversaries if they do not rapidly access and deploy technologies developed in the private sector.

The accelerating pace of innovation and the importance of private-sector R&D have strong implications for national security. Many of the technologies that are central to U.S. military predominance over peer and near-peer competitors—precision munitions, unmanned aerial systems, and other technology-enabled capabilities deployed in the past eighteen years—emerged from research sponsored by the federal government. The Pentagon funded almost 50 percent of the research and development in semiconductors from the 1950s until the 1970s. The shift to the private sector began in the 1980s, and by 1999 the Defense Science Board Task Force on Globalization and Security noted that the Defense Department was “relying increasingly on the U.S. commercial advanced technology sector to push the technological envelope and enable the Department to ‘run faster’ than its competitors.”⁴¹

Today, the Defense Department cannot remain ahead of potential adversaries without access to an expanded pool of technologies developed in the private sector. Technology companies innovate the software, computational capabilities, data analytics, and processing speed that drive the leading edge of cyber, space-based, unmanned, autonomous, and other complex military systems. While the DOD’s research, development, test, and evaluation (RDT&E) budget increased from $37 billion to $66 billion, or roughly 175 percent, over the past four decades, global R&D spending increased more than 1,875 percent during a similar period.⁴² In 2015, the top four U.S. defense contractors combined spent only 27 percent of what Google does annually on R&D.⁴³ This gap has only widened (see figure below).

Many of the tech sector’s new advanced technologies are multiple-use, benefiting both economic and military power. There are interconnections between the private sector’s investing in and developing these technologies and the military’s using them, but the private sector and military are governed by different rules and frameworks. Moreover, the defense base is national, while technology companies operate globally. The Defense Department has no monopoly on new technologies, as they are increasingly available to all nations, including U.S. competitors. As a result, competitive advantage derives from identifying emerging technologies and fielding complex systems more quickly. The 2018 National Defense Strategy notes, “Success no longer goes to the country that develops a new fighting technology first, but rather to the one that better integrates it and adapts its way of fighting.”⁴⁴

The U.S. military services have difficulty capitalizing on new technological developments in the private sector because of inflexible capital allocation and acquisition processes. As the 2018 strategy points out, the Defense Department “is over-optimized for exceptional performance at the expense of providing timely decisions, policies, and capabilities to the warfighter.” The battle systems of the future are software intensive, but bureaucracies designed to prevent waste and corruption are poor matches with a software development process that is iterative and where the end product is often unknown.

An April 2019 study on software acquisition by the Defense Innovation Board (DIB), an independent advisory committee to the DOD, concludes that “a large amount of DOD’s software takes too long, costs too much, and is too brittle to be competitive in the long run.”⁴⁵ Most DOD software projects adopt a “waterfall” development process that involves forming requirements, taking bids, selecting contractors, and then executing programs so they meet the listed requirements. The whole process can take so long that when the software is eventually deployed, it no longer matches operational needs. In addition, the Pentagon runs many closed proprietary and legacy systems, such as the Strategic Automated Command and Control System, which runs on a 1970s-era IBM computer system, and the Computerized Movement Planning and Status System, which uses Windows 2008.⁴⁶ Systems like these are expensive to maintain and force operators and warfighters into costly, time-consuming work-arounds. The DIB report warns that if the Pentagon “does not take steps to modernize its software acquisition and development practices, we will no longer have the best military in the world, no matter how much we invest or how talented and dedicated our armed forces may be.”

The DOD is only now adapting to start-up businesses, which require faster decisions and smaller funding increments. The economic incentives for start-ups to partner with the government are, however, mixed. Entrepreneurs must navigate complex bureaucracies and contracting processes. Most venture-backed companies are expected to start earning revenue within eighteen months, but it can take the Defense Department two years to award a contract, followed by testing, approval, and prototyping. In that time, the original technology may have also changed. Early-stage ventures thus often focus on the commercial sector, with its more reliable revenue and shorter sales cycles.

Not everything, of course, can “fail fast.” Weapons platforms that involve large numbers of warfighters in the loop, such as airplanes, submarines, and ships, will always demand longer development times, exceptional performance, and steady oversight. The challenge for lawmakers and Pentagon leaders is to balance risk and supervision appropriately.

In-Q-Tel, a venture capital firm established by the Central Intelligence Agency, has been investing in start-ups with defense- and intelligence-related technologies since 1999. A number of more recent efforts have been made to shift portions of the Defense Department’s procurement and deployment process closer to the funding models of the tech sector. DARPA’s Cyber Fast Track program, now closed, opened up R&D competitions to start-ups and hackers, who rarely worked with the DOD. The program awarded 130 contracts at an average cost of nearly $150,000 between two and sixteen days after first proposal.⁴⁷ The Defense Innovation Unit (DIU), created in 2015, has opened offices in Austin, Boston, and Silicon Valley to develop relationships with technology ecosystems and streamline procurement processes. In its first years, DIU awarded sixty-five contracts worth $200 million in the areas of artificial intelligence, autonomy, human systems, information technology (IT), and space.⁴⁸ DIU has also fostered the growth of software and hardware companies seeking specifically to serve the Defense Department. These are welcome efforts that should now be bolstered and expanded.

A Shortage of Tech Talent in Defense

The defense community faces severe challenges in attracting and retaining tech talent.

The armed services and Pentagon face a severe shortfall of talent in software development. They also often lack commercial cloud computing, agile software development environments, common machine-learning platforms, and other digital infrastructure with which coders are accustomed to working.⁴⁹ For example, in 2017 the Air Force had approximately 400 enlisted airmen and zero officers formally coded in the software development career field, out of roughly 320,000 uniformed Air Force members. Technical specialization is rarely good for career advancement in the military. The 2019 DIB report on software acquisition notes that “talented software developers and acquisition personnel with software experience are often put in jobs that do not allow them to make use of those talents, particularly in the military where rotating job assignments may not recognize and reward the importance of software development experience.”

A number of programs have tried to develop new technology talent within the Defense Department. Kessel Run, an Air Force project supported by DIU and involving a partnership with Pivotal Labs, has trained seventy airmen in software and application development.⁵⁰ The Defense Digital Service (DDS) recruits individuals from private technology companies for a limited tour of duty with the Pentagon. Jyn Erso, run out of the DDS, pairs talent from the private sector with the army’s top technologists, including in Army Cyber Command.⁵¹ DDS is a small program, having grown from roughly a dozen staffers to close to seventy in 2019, and relies on temporary personnel. These programs are a good start, but much more is needed. The lack of a direct career path for those with technical skills in the services and the insistence on training talent internally rather than hiring top external experts remain major obstacles in recruiting and retaining talent.⁵²

The Challenges of Hardware

The defense community faces deteriorating manufacturing capabilities, insecure supply chains, and dependence on competitor nations for hardware.

In addition to its software and talent problems, the Defense Department also faces complications in hardware innovation. Many of the sectors that Chinese manufacturers hope to dominate, and that the Chinese government supports through industrial policies, are the same as the DOD’s research and engineering priorities. U.S. manufacturers, like their international competitors, create their most cutting-edge hardware products from complex supply chains that span the globe. Critical design functions and high-end components are generally produced in the more advanced economies of the United States, Germany, Japan, South Korea, Taiwan, and others, with most of the IP and talent remaining—and growing—in those locations. Much of the final assembly of the components happens in China. Beijing is actively seeking, both through its own companies and through interaction with global multinationals, to encourage more valuable parts of the value chain to occur within its borders, or, when they are overseas, through supply chains that allow China to control or access the most relevant intellectual property.

Skills have also been lost from the domestic workforce since 2011, when the Budget Control Act required sequestration of $109 billion, affecting both mandatory and discretionary spending. Defense spending has increased since then, but the lack of predictable funding during that period led to the loss of roughly seventeen thousand defense vendors. The overall result has been deteriorating manufacturing capabilities, insecure supply chains, and a high level of dependence on competitor nations.⁵⁴ Policymakers, in consultation with the private sector, need to develop a more sophisticated view of global supply chains. Specifically, for critical technologies they should delineate which parts of the value chain are the highest priority for the United States: which components should be manufactured within the United States, which are most important for its companies to lead regardless of location of activity, and which would be most troubling to cede to Chinese companies or international companies operating in China.

High-tech start-ups are unlikely to fill the hardware innovation gap on their own. Companies built around hardware face high risk in terms of technology development and high costs associated with building research facilities, attracting scientific expertise, and manufacturing. The average amount required for a first funding round (known as Series A) for hardware companies is between $5 and $20 million, and subsequent rounds can reach as high as $50 to $100 million. The average Series A investment in a software-based company is between $1 and $3 million. Given the smaller risks of investing in software, VC firms funnel the vast majority of their investments to software, resulting in a funding gap for hardware. In 2017, 92 percent of U.S. VC dollars—up from 55 percent in 2006—went toward software-based technologies that have lower capital requirements, less invention risk, and quicker returns. Unless support for hardware manufacturing increases, the United States will rely increasingly on foreign companies that produce abroad, including in China.

A Persistent Divide

A persistent cultural divide between the technology and policymaking communities threatens national security by making it more difficult for the Defense Department and intelligence community to acquire and adopt advanced technologies from the private sector and to draw on technical talent.

Addressing the Defense Department’s hardware, software, and talent shortcomings is made more difficult by a persistent cultural divide that has been deeply exacerbated by some genuine policy differences and the current domestic political environment. This fissure has historical precedent. Draper Laboratory was spun off from the Massachusetts Institute of Technology (MIT) and became an independent research organization after protests from students and scholars about research in support of the Vietnam War. The current divide emerged over which uses of surveillance and encryption are legitimate, with the technology companies often siding with global users’ privacy concerns over the needs of intelligence and law agencies.⁵⁵ The divide has widened as employees at several of the largest technology companies have protested the use of artificial intelligence, facial recognition, and other frontier technologies for defense, intelligence, and homeland security projects. Google, for example, issued a set of AI principles, and after protests from Google engineers, the company did not renew its work on Project Maven, an artificial intelligence project with the Pentagon. Google also withdrew a bid from the Defense Department’s Joint Enterprise Defense Infrastructure (JEDI) project, a $10 billion IT improvement program, because it said the project clashed with its AI principles.⁵⁶ These moves and their effect on relations between the national security and engineering communities were viewed by some in a more negative light because of reporting that the company had been developing Dragonfly, a search engine for the Chinese market.

In effect, Defense Department officials see the United States as in an emerging arms race with China over AI and other technologies and have expressed concern about the role U.S. companies are playing in Chinese technology development. General Joseph F. Dunford Jr., chairman of the Joint Chiefs of Staff, noted the troubling implications of the behavior of some U.S. companies: “I have a hard time with companies that are working very hard to engage in the market inside of China, and engaging in projects where intellectual property is shared with the Chinese, which is synonymous with sharing it with the Chinese military, and then don’t want to work for the U.S. military.”⁵⁷ U.S. companies have much more complex views that vary enormously by company. Most see China as a huge market opportunity critical to their evolving supply chains; but they also see Chinese firms as fierce competitors and state-sponsored threats to their intellectual property. The growing presence of Chinese firms in Silicon Valley has deepened these views.

Despite these public contretemps, a fair number of technologists are willing to work with the U.S. government. In a February 2019 BuzzFeed survey of one thousand tech workers in Silicon Valley, 59 percent of respondents “somewhat agree” or “strongly agree” that “tech companies should work with the U.S. government on military projects,” whereas only 31 percent of tech workers “somewhat agree” or “strongly agree” that U.S.-based tech companies should operate in China.⁵⁸ Top executives at Amazon and Microsoft have also affirmed their willingness to work on classified contracts for the military and the intelligence community, while others, including Apple, build products that are already in use there.⁵⁹

The Task Force believes that closing the divide between policymakers and the tech industry is essential to national security. It will require greater transparency, explanation, outreach, and experimentation by the Defense Department, as well as increased interaction between members of the military and technology communities. In addition, tech leaders should frequently and publicly explain to their employees that they have the best chance of shaping the development and use of frontier technologies by working with the Pentagon directly.

China is investing significant resources in developing new technologies, and after 2030 it will likely be the world’s largest spender on research and development. Although Beijing’s efforts to become a scientific power could help drive global growth and prosperity, and both the United States and China have benefited from bilateral investment and trade, Chinese theft of intellectual property and its market-manipulating industrial policies threaten U.S. economic competitiveness and national security.

The pressure on the United States is heightened by the rise of other science and technology competitors, especially China. Countries in Asia and Europe have steadily improved their innovation ecosystems, rolling out R&D tax incentives and increasing government funding for research and technology commercialization initiatives. In 1960, the United States accounted for 69 percent of global R&D. By 2018, the United States’ share had fallen to a little over 25 percent, with 43.6 percent of the spending emanating from Asia.⁶⁰ Japan, Singapore, South Korea, and Taiwan have all seen science and technology as essential to economic security. Seoul, for example, increased spending on R&D as a percentage of GDP from 2.1 percent in 2000 to 4.5 percent in 2017.⁶¹

China in particular has ambitious plans to become a world leader in science, technology, and medicine. Between 1991 and 2015, China increased its R&D expenditures thirtyfold, averaging an 18 percent increase annually since 2000.⁶² In nominal terms, Chinese R&D expenditures rose to $254 billion in 2017, approximately 45 percent of U.S. R&D spending for that year. Adjusted for purchasing power, China’s R&D expenditures were closer to 88 percent of U.S. spending.⁶³ China’s GDP is growing and China is dedicating a greater portion of its economic resources to R&D, planning to eventually reach a spending target of 2.5 percent of GDP. It will likely equal or exceed the United States in overall R&D expenditures after 2030 (see figure below).

The STEM workforce in China has also rapidly expanded. The total number of Chinese universities grew from 1,792 to 2,560 between 2005 and 2015. Eight million Chinese students graduated from college in 2017, compared to approximately 1.9 million graduating with bachelor’s degrees and 1 million with associate’s degrees in the United States.⁶⁴ The number of science and engineering bachelor’s degrees conferred in China increased from 359,000 in 2000 to 1.65 million in 2014.⁶⁵ China surpassed the United States as the world’s largest producer of natural sciences and engineering doctorates in 2007 (see figure below).⁶⁶ Questions have been raised about the quality of some Chinese programs, but there is no doubt that the Chinese ability to compete in STEM fields has grown.

In addition, with ambitious science projects, generous salaries, and high levels of lab funding, China has made a concerted effort to recruit top foreign talent.⁶⁷ The Thousand Talents Program offers scientists a one-million-yuan ($151,000) starting bonus and research funds of three to five million yuan. Foreign scientists receive additional incentives, such as subsidies for accommodation, visits home, and education.⁶⁸ The Department of Energy recently warned that talent programs were offering scientists at U.S. national labs hundreds of thousands, and in some cases millions, of dollars to conduct research in China.⁶⁹

China is closing the technological gap with the United States, and though it may not match U.S. capabilities across the board, it will soon be one of the leading powers in technologies such as AI, robotics, energy storage, 5G, quantum information systems, and possibly biotechnology.

This increase in spending and STEM personnel is beginning to pay off in scientific accomplishments. China overtook the United States in the production of scientific papers in 2016. According to a study by scientific publisher Elsevier and business news outlet Nikkei, China published more high-impact research papers than the United States did in twenty-three out of thirty research fields with clear technological applications. China has built many of the world’s fastest supercomputers and will likely beat the United States to building the first exascale supercomputer, despite investing around the same amount in supercomputers as the United States.⁷⁰ China’s current five-year plan prescribes that the biotechnology sector should exceed 4 percent of GDP by 2020, and state, provincial, and local governments have invested more than $100 billion in the life sciences sector.⁷¹ BGI (formerly the Beijing Genomic Institute) is by some measures the largest genome-sequencing center in the world, and an increasing number of U.S. companies depend on Chinese partners to do their sequencing and analysis. China also plans to build the largest-ever particle accelerator. In 2016 it became the first country to send a quantum-encrypted message via satellite, and in 2018 it held the first quantum-encrypted video call. In January 2019, China became the first country to land a vehicle on the far side of the moon.

China has three industrial policies designed to raise its innovation capabilities: the 2014 Integrated Circuit (IC) Promotion Guidelines, Made in China 2025, and the Next-Generation Artificial Intelligence Development Plan. The IC Promotion Guidelines, an attempt to build an indigenous integrated circuit industry, involves investments reportedly between $100 and $150 billion in public and private funds. The goal is to have Chinese firms produce 70 percent of the chips consumed by Chinese industry, reducing their dependence on U.S., Korean, and Taiwanese suppliers. Made in China 2025 sets ambitious targets for upgrading China’s aging manufacturing base through smart manufacturing and offers low-interest loans from state-owned investment funds and development banks, assistance in buying foreign competitors, and extensive research subsidies.

On AI, Beijing hopes to leverage massive amounts of data, permissive regulations, entrepreneurial firms, and government support to build an industry worth $150 billion by 2030. In 2017, China’s AI industry received nearly $26 billion in investment and financing.⁷² The United States still leads in cutting-edge R&D, specialized chips, and talent, but China surpassed the United States in volume of AI research in 2014, including in AI-related patent registration and articles on deep learning. China is also training a large number of specialists.⁷³ Twenty-three percent of the accepted papers for the 2017 Association for the Advancement of Artificial Intelligence conference were from China, rising from only 10 percent in 2012, and AI authors in China were cited 44 percent more in 2016 than they were in 2000 (see figure below).⁷⁴ China will open around four hundred majors related to data science, artificial intelligence, and robotics in universities in 2019.⁷⁵

Although Chinese companies have smaller R&D budgets than their American competitors, they are world leaders in a number of frontier technologies. The country’s two largest internet companies—Alibaba and Tencent—have developed highly innovative e-commerce and mobile payment platforms. Datang, Huawei, and ZTE own about 10 percent of 5G-essential intellectual property rights (IPR), and many analysts expect China to fully commercialize 5G by 2020, five years ahead of the United States, Australia, the EU, Japan, and South Korea (see figure below). Officials at Huawei announced that they planned to more than double annual R&D spending to between $15 billion and $20 billion, which would place the company between second and fifth place in worldwide spending on R&D.⁷⁶ Chinese companies such as Baidu, ByteDance, Face++, iFLYTEK, and SenseTime are driving the application of AI to voice and facial recognition software, autonomous vehicles, and internet content. In the first and second quarters of 2018, the volume of Chinese VC investment surpassed that of the United States for the first time.⁷⁷

Civil-military fusion is a pillar of Chinese military modernization and an effort to bolster the country’s innovation system for advanced multiuse technologies in aviation, aerospace, and information technology. Introduced by then President Hu Jintao in 2009, the effort to bridge the gap between the civilian industrial base and the military has intensified under President Xi Jinping. Within his first year in office, the Central Committee voted to elevate civil-military fusion to a national strategy, and in January 2017 Xi created the Central Commission for Integrated Military and Civilian Development, a high-level decision-making and coordination body for civil-military fusion efforts. The top seven state funds dedicated to investing in civil-military fusion industries report having over 362 billion yuan ($56.85 billion) in capital.⁷⁸ Civil-military fusion plays a prominent role in both Made in China 2025 and the Next-Generation Artificial Intelligence Development Plan. Beijing’s policies could help it achieve a meaningful edge in the development of future weapons systems, reducing or perhaps eliminating a longstanding source of strategic advantage for the United States.

China also seeks to influence the digital infrastructure of the future through the Belt and Road Initiative. Along with investment in railways, roads, pipelines, and ports along the route, Chinese companies plan to build a “digital Silk Road”: fiber-optic cables, mobile networks, satellite relay stations, data centers, and smart cities. ZTE, for example, operates in over fifty of the sixty-four countries on the route of the Belt and Road Initiative. It lays fiber-optic cables; sets up mobile networks; and provides surveillance, mapping, cloud storage, and data analysis services to cities in Ethiopia, Laos, Nigeria, Sri Lanka, Sudan, and Turkey.⁷⁹ Beijing hopes these economic ties will translate into political influence over the shape of the internet and the rules governing emerging technologies.

The Chinese innovation model is not without its own weaknesses. China spends only about 5 percent of its R&D funds on basic research, compared to 17 percent in the United States.⁸⁰ Chinese firms depend on American technology in some critical areas, especially semiconductors. Top-down direction and industrial policy often leads to waste, corruption, and redundancies. Plagiarism and fabrication of scientific results are perennial problems at Chinese universities and research labs. Although an increasing number of foreign-trained Chinese scientists are returning home, they often find that the research environment is hierarchical and grants depend on political and personal connections. Moreover, the Chinese Communist Party is now reasserting political control over technology companies, which for the last decade benefited from a relatively laissez-faire environment. This may slow the introduction of new products and innovation.

The Chinese leadership is aware of these barriers to innovation and is beginning to address them. Chinese officials have also interpreted U.S. efforts to cut off the flow of technology to Huawei and other companies as an effort to contain China’s rise as a science and technology power. They have responded by doubling down on innovation and self-reliance, with the objective of reducing their dependence on U.S. technology, especially in the semiconductor and semiconductor-tool industries.⁸¹

The challenge from Beijing is pressing and immediate, but the United States needs to look beyond China, and beyond competition over a specific list of cutting-edge technologies. There is much that Washington can and should do that is unrelated to Beijing and is instead concentrated on maintaining U.S. leadership. The U.S. government, the private sector, and academia should work together to increase the national capacity for scientific and technological innovation and accelerate the adoption and deployment of new technologies by defense and intelligence agencies.

Although the Trump administration has boosted the budgets of several technology-related organizations within the DOD and issued a number of executive orders, its efforts to accelerate innovation in critical frontier technologies such as AI are too incremental and narrow in scale.

The Task Force commends the Trump administration for bringing much-needed attention to innovation issues and for shining a spotlight on the development of critical technologies such as 5G and AI. However, the White House has not taken on the challenges in a comprehensive way that will produce durable results. Despite bipartisan support for broad technology competition with Beijing, the White House has failed to work with Congress to increase federal support for basic R&D and has adopted an incremental and limited approach to supporting the development of frontier technologies. The White House’s immigration policies have weakened the country’s ability to compete for talent, and unnecessary trade conflicts with friends and allies have hampered the building of international technology coalitions and could slow innovation.

The Trump administration’s innovation strategy has combined a commitment to deregulation with a number of executive orders in support of frontier technologies, the expansion of broadband, and workforce development.⁸² Administration officials have stressed that while the federal government has a role to play in making important investments in R&D, it is the private sector that drives innovation, and thus a large focus should be on removing regulatory barriers that slow American entrepreneurs. In March 2017, Trump established the White House Office of American Innovation (OAI), which aims to streamline government and cut red tape. The office’s remit is relatively narrow, given the scope of the innovation challenge, and in its first years it focused on two information technology initiatives: publishing a report on IT modernization and creating the Centers of Excellence program to drive agencies to the cloud and better data management.

The Trump administration has, however, also pushed for a more muscular government response in areas where it sees innovation as a matter of great-power competition. The White House’s FY 2019 budget called for increasing funding for RDT&E of new technologies in the Defense Department to its highest level since the end of the Cold War. The majority of this funding increase would support late-stage development, prototyping, and testing activities.⁸³ The administration requested a $360 million increase to DARPA’s funding in FY 2019, to $3.432 billion from $3.07 billion in FY 2018, and an additional $125 million, to $3.556 billion, in FY 2020.⁸⁴ The Defense Innovation Unit saw a nearly trifold increase in its budget under the administration’s FY 2019 budget, from $29.6 million to $71 million, which then more than doubled to $164 million in FY 2020.⁸⁵ The Task Force strongly encourages these important signals of continued support for two innovative organizations within the DOD, but the scale remains small compared to the massive Defense Department budget.

The Trump administration has also increased its focus on cutting-edge technologies such as semiconductors, quantum computing, and artificial intelligence. In 2017, DARPA launched the $1.5 billion, five-year Electronics Resurgence Initiative to support work on advanced chip design and manufacturing. In September 2018, the White House held the Summit on Advancing American Leadership in Quantum Information Science, at which the Department of Energy announced $218 million in funding and the National Science Foundation $31 million to support multidisciplinary quantum research. Trump signed the National Quantum Initiative Act in December 2018, which authorizes the government to provide $1.2 billion to fund activities promoting quantum information science over an initial five-year period. These are all positive steps, but they are incremental and fairly limited in scope. To truly meet the innovation challenge, a much broader and more sustained approach on multiple fronts—including federal investment in R&D, a moonshot approach to innovation policy, comprehensive immigration reform, and partnerships with friends and allies—will be needed.

Artificial Intelligence

The United States is ahead of the rest of world in AI, but others are closing the gap—and U.S. failure to compete for global talent could result in the loss of its lead.

The Task Force is concerned that the White House has been slow in driving the development of AI, failing to build immediately on the Barack Obama administration’s release of a national AI plan in 2016.⁸⁶ A growing number of governments have introduced national strategies: Canada, China, Denmark, the EU Commission, Finland, France, Germany, India, Italy, Japan, Mexico, the Nordic-Baltic region, Singapore, South Korea, Sweden, Taiwan, the United Arab Emirates, and the United Kingdom have all released strategies to promote the use and development of AI.

In response to this criticism, officials note a steady stream of actions on AI during 2018 and 2019. The FY 2019 National Defense Authorization Act (NDAA) created the National Security Commission on AI, which has been tasked with assessing the national security implications of AI. In June 2018, the Pentagon announced the Joint Artificial Intelligence Center (JAIC), which will partner with industry and academia to help the DOD utilize AI. The FY 2019 NDAA includes $70 million in funding for the center.⁸⁷ DARPA has committed to a multiyear $2 billion investment in new and existing programs, in its AI Next campaigns.⁸⁸

The White House also held a summit in May 2018 with industry representatives to discuss AI policies and announced that federal agencies would make data available for private-sector AI research.⁸⁹ It established a new Select Committee on AI, housed under the National Science and Technology Council and including representatives from DARPA, the NSF, and the OSTP. The committee will advise the president on government priorities in AI and build private-sector partnerships.

On February 11, 2019, Trump signed an executive order enabling the American AI Initiative.⁹⁰ A day later, the Defense Department released its AI strategy, designed to accelerate the delivery and adoption of AI; strengthen partnerships with industry, academia, allies, and partners; cultivate an AI workforce; and lead in military AI ethics and safety.⁹¹ Along with the executive order, the White House issued a National Security Presidential Memorandum, “Protecting the United States’ Advantage in Artificial Intelligence and Related Critical Technologies,” which charges the assistant to the president for national security affairs with coming up with a plan to protect the United States’ advantage in AI from strategic competitors and foreign adversaries.

These efforts, however, are inadequate. Action does not match the language officials use to describe the importance of AI to U.S. economic and national security. The American AI Initiative provides no new sources of federal support and is unclear about how much the government should be spending on AI research and development. Individual agencies are told to “budget an amount for AI R&D that is appropriate for this prioritization,” but without additional funding, they will need to shift funds, leaving other research areas underfunded. Lacking any specific spending goals, the initiative has no ability to measure whether an agency is adequately prioritizing AI research. Few details were given for the initiative’s execution.⁹²

Huawei and 5G

In the race for the next generation of communications technologies, the Trump administration has developed only a few parts of what should be a multifaceted strategy. It has failed to coordinate a response to Huawei’s global expansion, muddied its message about the company’s economic and national security risks, and not sufficiently accelerated domestic efforts to deploy 5G.

5G will offer data speeds up to fifty or one hundred times faster than current telecom networks and will serve as critical infrastructure for AI, automated vehicles, the Internet of Things, and other industrial sectors. In October 2018, Trump signed a memorandum directing the Commerce Department to develop a long-term, comprehensive national strategy for spectrum (that is, radio waves, which are used in telecommunications) and ordering federal agencies to review their existing spectrum usage, forecast future demands, and prepare a plan for research and development that will enable better use of spectrum in the future.⁹³ In April 2019, the Federal Communications Commission (FCC) announced the third, and largest, auction of high-frequency spectrum for 5G—opening up use of high-frequency spectrum to companies so they can roll out 5G commercially—and a $20 billion fund to expand broadband in rural areas.⁹⁴

In addition, the Trump administration has moved to block Chinese telecom firms from rolling out 5G in the United States and foreign markets. In May 2019, Trump signed two executive orders to that effect. The first authorized the Commerce Department to block U.S. companies from using telecom equipment and services from companies controlled by “adversary governments.” U.S. officials have claimed the order is “agnostic,” but it has been widely interpreted as being directed at China and Huawei.

The second order has larger consequences for the company. It ordered the Commerce Department to place Huawei and sixty-eight affiliates on a list of companies to which U.S. firms may not sell components without government approval. Broadcom, Intel, Qualcomm, and Xilinx stopped working with Huawei after the order was announced, and Google stated that it would no longer provide the Android mobile operating system and apps for Huawei’s smartphones. Cut off from U.S. chip and software suppliers, Huawei’s ability to operate—and its future—were highly uncertain. Huawei founder Ren Zhengfei has said he expected company revenues to decline by $30 billion over the next two years because of U.S. actions.⁹⁵ During the June 2019 Group of Twenty summit in Japan, however, Trump agreed to lift some of the sanctions against Huawei, allowing sales of widely available components made by American companies.⁹⁶

The White House and the State Department have also, with mixed results, tried to convince other countries not to use Huawei and other Chinese telecommunications equipment in their next-generation wireless networks. While Australia, Japan, and New Zealand have banned Huawei, other friends and allies are moving ahead, despite threats from the United States to limit intelligence sharing. Most notably, officials in France, Germany, and the United Kingdom have argued that they can manage security risks by developing strict standards, inspecting equipment and code, and installing Huawei equipment only on peripheral, controlled networks.⁹⁷

The Task Force believes the White House was right to publicize the security risks of Huawei and block adoption of the company’s equipment in U.S. networks. The use of the Commerce Department list to ban sales to Huawei, however, was too blunt an instrument and caused significant blowback for U.S. technology companies. It has also encouraged China, and others, to reduce dependence in areas where American technology companies dominate, such as semiconductors and design tools.

Moreover, Trump’s willingness to overturn the sanctions on Huawei suggests that the sanctions were based on U.S. economic interests rather than legitimate security risks, which undercuts efforts to convince other countries to exclude the company. Senator Marco Rubio (R-FL) tweeted that a reversal on Huawei would be “a catastrophic mistake” that would “destroy the credibility of [the Trump] administrations warnings about the threat posed by the company, no one will ever again take them seriously.”⁹⁸

Technology Protections

Beijing has often exploited the openness of the American system. Efforts to protect U.S. intellectual property are a necessary complement to, but not a substitute for, innovating faster than China. The administration is over-weaponizing trade and investment policy, with costs to U.S. innovation.

The campaign against Huawei has at times overlapped with a major facet of the Trump administration’s innovation strategy: protecting American technology at home and abroad. In August 2018, Congress passed the Foreign Investment Risk Review Modernization Act (FIRRMA), which broadens the jurisdiction of the Committee on Foreign Investment in the United States (CFIUS), allowing it to investigate, and possibly block, more foreign deals. The Committee may now investigate a foreign entity buying not just an entire company but also minority, noncontrolling investments. The Trump administration has blocked the sale of Lattice Semiconductor to a group that involved a Chinese venture capital firm; barred Broadcom’s $121 billion offer for Qualcomm; prevented Ant Financial’s acquisition of MoneyGram; and demanded that Beijing Kunlun Tech give up control of Grindr.

Total Chinese direct investment in the United States has fallen 90 percent, to $5 billion last year from $46 billion in 2016, driven in part by FIRRMA but mainly by tighter controls on outward investment from Beijing.⁹⁹ The legislation is beginning to affect early-stage investments: some capital has moved into new sectors that are not as politically sensitive, and some dealmakers in Silicon Valley say Chinese funds are looking for deals outside the United States to avoid scrutiny.¹⁰⁰ American venture capital firms are reportedly dropping their Chinese investors or walling them off, and some start-ups have forced out Chinese investors to avoid regulators.¹⁰¹ In April 2019, PatientsLikeMe, a health-care start-up, was ordered to find a new buyer after the Trump administration forced its Chinese majority shareholder to divest its stake.¹⁰² Once the new rules are fully implemented, the drop-off could be even more noticeable.¹⁰³

It is not only the inflow of money that has provoked security concerns. The size of the Chinese student population in the United States—an estimated 350,000, about half of whom are studying at the undergraduate or lower levels—presents a challenge to law enforcement and counterintelligence agencies. FBI Director Christopher Wray and other U.S. officials have recently warned that Chinese intelligence is using expatriate scientists and students to gain access to technologies at universities and businesses.¹⁰⁴ The FBI, federal granting agencies, and members of Congress have signaled that universities need to do more to prevent foreign actors from attempting to steal intellectual property.¹⁰⁵ In June 2018, the State Department implemented a one-year limit on visas for Chinese graduate students studying in sensitive research fields, with the chance to reapply every year. The administration is reportedly considering new background checks and other restrictions on Chinese students.¹⁰⁶

The White House is also deploying new export controls to slow the pace of Chinese development. The Commerce Department is developing regulations to restrict “emerging and foundational technologies,” including robotics, 3-D printing, and biotechnology, as well as several categories of AI, including computer vision, speech recognition, and natural language understanding.¹⁰⁷ In June 2019, the Commerce Department prohibited U.S. chip companies from selling to five Chinese entities involved in developing exascale computing, including the supercomputer maker Sugon.¹⁰⁸

The United States is also trying to pressure Beijing into ending the theft of IP and trade secrets from U.S. companies. Presidents Xi and Obama had signed an agreement in 2015 to refrain from economic espionage, but Chinese hackers have returned and targeted numerous corporations, including cloud providers and IT service suppliers.¹⁰⁹ In November 2018, then Attorney General Jeff Sessions announced the China Initiative, to identify priority trade-theft cases, pool Department of Justice and FBI resources to combat Chinese economic espionage, and evaluate whether additional legislative and administrative authorities would be required to protect U.S. assets from foreign economic espionage.¹¹⁰ Between October and December 2018, the Department of Justice unsealed indictments three times against Chinese intelligence officers and hackers for the theft of U.S. businesses’ IP and trade secrets.¹¹¹

The Task Force commends the White House for confronting China on cyber espionage and IP theft. Updating CFIUS and export controls is also overdue, especially in the case of Sugon, which has connections to China’s People’s Liberation Army. The Task Force warns, however, that the administration is over-weaponizing trade policy, with long-term costs to U.S. innovation capabilities. The issue is not only the loss of revenues to U.S. tech companies from Chinese customers, though these are significant. It is also that Beijing and others will want to reduce dependence on U.S. high-tech supply chains now that they have seen them leveraged for political goals.

The Task Force believes investment restrictions and export controls are a necessary but secondary part of any strategy responding to China’s rise as a science and technology power. As such, limitations on the flow of people and money should be drawn as narrowly as possible. Slowing China down is not as effective as outpacing it. The United States needs its own innovation policies.