Table of Contents

Risks of Interdependence

As China’s technological prowess grows, U.S. officials worry that China could come to lead and possibly dominate the most economically significant tech industries of the future. This concern is not about China’s unfair practices per se; it is about unfavorable outcomes for U.S. competitiveness. The distinction is often glossed over, yet it is crucial. Even on a level playing field, China could potentially outcompete the United States in some tech industries. This troubles American policymakers, most of whom have only known an era of peerless technological leadership by U.S. companies. They now see China catching up (or even moving ahead) in the technology areas expected to matter most for twenty-first-century economies.

China has already become the global leader in 5G telecommunications equipment (a crux of the future digital backbone), as well as commercial drones, Internet of Things devices, mobile payments, solar cells, and smart cities, among other technology areas. And where China does not lead, it is often a world-class competitor—for example, in AI (the most-hyped of all emerging technologies), smartphones, electric vehicles, and much more. While unfair economic practices have contributed to China’s success, they do not tell the whole story. China ranks first globally in STEM graduates and second in R&D spending.1 Its geographically concentrated tech hubs have revolutionized supply chain and manufacturing integration. And government tech policies like Made in China 2025 and the Digital Silk Road, although often misunderstood and overestimated in the West, nevertheless demonstrate Beijing’s national focus on technology strategy—a relative weak point for the United States in recent decades.2

If China halted all unfair practices tomorrow, its tech industry would still likely represent the most significant challenge to U.S. technology leadership and global competitiveness since the rise of Japan in the 1980s.3 Of course, China’s challenge might fade over time, just as Japan’s did. China could experience a bursting debt bubble, or a “middle income trap,” among other potential causes of technological stagnation. But Washington has little ability to predict, let alone influence, such developments. If the current trajectory continues, the United States will have several problems on its hands.

If China halted all unfair practices tomorrow, it would still likely represent the most significant challenge to U.S. technology leadership since Japan in the 1980s.

At the most basic level, losing America’s technological edge in major industries would mean fewer U.S. jobs, lower GDP, reduced tax revenue, and other macroeconomic setbacks. It would also diminish the global influence that America derives from its technology leadership. For example, U.S. dominance of digital platforms has provided unparalleled intelligence collection opportunities and helped to project certain American political and cultural values into foreign societies. It also provided the Biden administration with concrete leverage to shape and secure a recent global tax agreement.4 Loss of American technological dominance would lessen those forms of power and influence. Moreover, it would weaken the so-called “national security innovation base,” a Washington term for the American tech industry’s special role in generating new U.S. military and intelligence capabilities.5 No country wants to fall behind in these ways, and the United States is particularly reliant on its technological leadership as a source of economic and national security advantages.

Still, the United States could survive and even thrive despite a loss of dominance in some important tech areas—so long as it remains relatively competitive overall. The more serious threat is that China itself becomes so technologically dominant that American companies are largely frozen out of many important markets. U.S. leaders worry that China’s technology gains could become strongly self-reinforcing, paving the way toward just this kind of dominance.6 Certain tech markets—5G telecommunications equipment is a prime example—have high barriers to entry, significant first-mover advantages, and deep linkages to many other sectors. In such cases, China’s early leadership could enable its companies to lock in global market share and seek to dominate related or adjacent industries. Once Chinese firms secured strong enough positions, they and the Chinese government could use unfair practices, like predatory pricing and exclusionary deals, to further entrench their advantages.

These worries have driven U.S. policymakers to try to curb bilateral tech ties that Washington sees as helping China catch up with or overtake the United States—especially in the most economically important technology areas, where long-term leadership is at stake. Two laws passed in 2018 illustrate this trend.

The Foreign Investment Risk Review Modernization Act (FIRRMA) calls for CFIUS to scrutinize transactions involving “a country of special concern that has a demonstrated or declared strategic goal of acquiring a type of critical technology or critical infrastructure that would affect United States leadership in areas related to national security.”7 This is a clear mandate to thwart Made in China 2025, Beijing’s strategic technology development plan. Likewise, the Export Control Reform Act (ECRA) instructs the Commerce Department to limit exports of “emerging and foundational technologies.” Although these categories remain undefined, Made in China 2025 has served as a starting point for U.S. government analysis.8 Technically, both FIRRMA and ECRA describe national security as their sole focus. But economic considerations helped spur Congress to act and will certainly influence future regulatory actions.

Risks and Limitations of Defensive Measures

Preventing China from seizing control of key tech industries is a worthy policy goal. The United States should not underwrite its own economic dislocation if it can avoid doing so. Nevertheless, there is risk of overreach—particularly if Washington views technology controls as the primary means of maintaining its own preeminence, rather than as stopgap measures to thwart Chinese dominance so that other American investments have time to take root. U.S. policymakers should keep several points in mind.

First, the “cure” of government technology controls can sometimes be as harmful as the “disease” of aggressive Chinese competition. Export controls, Entity List designations, and similar restrictions reduce American companies’ sales to China, cutting the revenue available to plow back into R&D. Visa bans, deemed export curbs, and supply chain security requirements restrict U.S. access to Chinese talent and subcomponents, imposing higher costs and greater delays on American innovators. Inbound investment restrictions limit U.S. firms’ opportunities to raise capital from and achieve corporate synergies with Chinese entities. If U.S. controls are unilateral, then European and Asian rivals can gain competitive advantages over American firms. Moreover, China can and will retaliate against significant restrictive measures.

For these reasons, trade groups like the Semiconductor Industry Association, the Business Roundtable, the National Association of Manufacturers, and the Information Technology Industry Council have sought to limit the use of ECRA and similar new authorities.9 Although trade associations have narrow vested interests that can diverge considerably from the national interest, the coalitions raising these concerns have been notably broad-based and diverse. They include a set of groups that represent “all major research universities and medical schools in the United States,” which collectively warned that “overly broad or vague controls will result in unnecessary restrictions that will stifle scientific progress and impede research.”10

Second, technology controls can conflict with Washington’s stated objective of fair, rules-based economic competition, as described earlier. WTO rules do not allow a country to curb trade in certain industries just because they are seen as economically important. In fact, this runs directly counter to WTO principles. If the United States publicly embraces an economic strategy of restricting trade in important tech industries, then China and other countries will step up protections of their own “strategic” industries (whether tech or non-tech). Although the WTO system urgently needs reforms, U.S. leaders have not yet articulated a credible vision for reform or a plan for gaining agreement from China and other major trading partners.

Third, it is not easy to identify economically strategic technologies that merit governmental controls. The U.S. government has historically struggled to make accurate, useful predictions about what innovations will be important in the future and to draw administrable lines around fuzzy technology areas.

In the 1990s, economic competition from Japan spurred the U.S. government and outside groups to produce a number of so-called critical technology lists to help inform policymaking.11 Defining “criticality” proved to be a major challenge, given the many different U.S. interests impacted by technology. In part for this reason, the lists were usually too broad (naming whole fields of practice, such as “programming languages”) and/or too long (some had over 100 items) to be useful in policymaking. In hindsight, it is also apparent that many items designated as critical turned out not to be, while some technologies excluded from the lists wound up having vast economic impact. For example, the 1995 White House list cited “virtual reality software” (still a marginal industry today) as critical, but omitted personal mobile devices (a then-extant technology that has since revolutionized global communications).12 Critical technology lists ultimately failed to have much policy impact, and the effort was abandoned in the 2000s.

The Trump administration sought to revive this moribund tradition by publishing a “critical and emerging technologies list” in October 2020.13 Unfortunately, this list reflected many old pitfalls. It was vague and equivocal in defining the criteria for inclusion—at one point saying the listed technologies were critical to overall U.S. national security and/or economic advantages, then later calling them critical merely for U.S. government agencies. Although the list was manageable in size (just twenty items), the individual entries had sweeping scope: “Energy Technologies,” “Communication and Networking Technologies,” “Data Science and Storage,” “Medical and Public Health Technologies,” “Biotechnologies,” and so on. And its publication so late in a presidential administration meant that Trump’s list, like others before it, lacked clear policymaking relevance.

The U.S. government should develop more detailed and robust internal processes for evaluating the economic consequences of emerging technologies.

The Biden administration refreshed this list in February 2022.14 The new categories of critical and emerging technologies do not differ much from Trump’s, though Biden’s version helpfully elaborates on each category by defining multiple subcategories. Still, the list’s selection criteria and policymaking purpose remain unclear. Does it name technologies that “may be critical to U.S. national security,” or merely those with “the potential to further” it? Both formulations are given. National security, meanwhile, is defined as including “economic prosperity and opportunity”—a defensible framing that nevertheless clouds the specific rationale behind each entry’s inclusion, making it harder to use the list as policy guidance. In fact, the White House emphasized that “this list should not be interpreted as a priority list for either policy development or funding.”15 Yet it also recommended that agencies consult the list when designing “measures that respond to threats against U.S. security” and “initiatives to research and develop technologies”—that is, policy development and funding.

Washington does not necessarily need (and probably shouldn’t try to create) a singular list of all technologies critical for every U.S. national interest. But it does need something beyond what it has now. Routine policy actions like export control listings and CFIUS investigations already require some predictions of future technological importance, despite the inherent difficulties. Yet formally, these policy tools have an exclusive focus on national security concerns, with economic interests left as a potential but largely undefined (and sometimes unspoken) consideration. The U.S. government should develop more detailed and robust internal processes for evaluating the economic consequences of emerging technologies, so that agencies can tailor controls to critical areas where China threatens to secure dominance.

Recommended Policies and Processes

Government technology controls should play a specific, limited role in the U.S. quest to maintain economic leadership and competitiveness. The president should instruct regulatory agencies to institute restrictive measures only when necessary to hold off looming Chinese dominance in defined strategic industries, thus buying time for other positive American investments to bear fruit. Implementing this guidance would involve creating a formal governmental process to conduct (or oversee) geo-economic analysis of technologies. A new process would need to avoid replicating 1990s-era mistakes while also adapting to today’s geostrategic, technological, and governance realities. Fortunately, scholars and independent analysts have offered a number of useful recommendations and lessons for policymakers.16

One approach would be for geo-economic assessments to identify technologies that rate highly across three dimensions: economic value, defensibility, and urgency of control. Each of these dimensions can be defined and measured in various ways. For example, highly economically valuable technologies might include those set to become top exports (as semiconductors are now), to produce the largest companies of the future (as with today’s digital advertising sector), or to have powerful second- or third-order effects on many other industries (like clean energy or advanced batteries).

Highly defensible technologies would have strong winner-take-all qualities, meaning that a market leader could capture disproportionate gains and then defend its position for long periods, perhaps due to network effects (as in today’s social media market) or high barriers to entry (as in telecommunications equipment). Frameworks for evaluating technological defensibility can be found in antitrust economics and venture capital investing, among other domains. Finally, technologies in urgent need of control are those where a window of competitive opportunity could soon close. Washington might assess whether technological and market developments during the next five to ten years could enable China to achieve and lock in long-term dominance.

If the United States publicly announces an official process for identifying and controlling economically strategic technologies, China and other countries would likely accuse the United States of a flagrant assault on WTO principles. Therefore, the U.S. government should seek to maintain its tradition of requiring that new technology controls have a national security justification, even if, internally, the initial impetus for considering a control is economic competition. Ideally, new controls should be framed as continuations of historic U.S. policy, to include American claims of a WTO national security exception; new precedents should be created only when necessary. The U.S. framework for identifying economically strategic technologies should probably remain confidential or classified.

Case Studies

5G. The Trump administration was right to identify 5G telecommunications equipment as a major target for technology restrictions. Countries all over the world are making or planning massive investments in such equipment, and deployment is expected to drive many ancillary innovations in areas such as autonomous vehicles, the Internet of Things, and mobile apps. Because these are generational investments, market leaders like Huawei have an opportunity to establish firm footholds in the purchasing countries. And the purchases are all happening within a short time frame as countries race to deploy 5G, meaning the window of opportunity will close within the next few years.17

Semiconductors. Semiconductors are also a strategic industry: they rank among the most important technologies by overall sales and are essential to almost all modern activity, while the most advanced semiconductors will drive next-generation technological applications like machine learning. Yet U.S. semiconductor firms also rely on China for much of their revenue, and most experts believe China still remains about a decade behind the United States and its allies in key aspects of semiconductor design and manufacturing.18 U.S. government controls on this sector should therefore be carefully targeted.

The United States should restrict China from accessing only the most advanced semiconductor technology, while allowing sales of commodity chips to help maintain U.S. market share and fund R&D. The Trump administration was right to press the Netherlands to prevent export of extreme ultraviolet lithography systems to China.19 These would facilitate manufacture of 5- and 7-nanometer node chips and thus help China to leap well ahead of its current capabilities. At the same time, the Biden administration was also right to grant U.S. firms licenses to sell automotive chips—considered commodity items—to Huawei.20 A more difficult case was Trump’s tightening of rules for so-called deemed exports, in effect requiring more scrutiny for Chinese nationals working in the U.S. semiconductor industry.21 Although the costs and benefits are difficult to independently assess, it makes sense in principle to prioritize protection of critical, cutting-edge intellectual property and trade secrets in the semiconductor sector.

Consumer devices. However, very few technologies sold in large quantities to individual consumers should be subject to government control on the grounds of economic criticality. Smartphones, laptops and desktop computers, Internet of Things devices, consumer-grade drones, home network hardware, gaming systems, and most mobile apps should be relatively unrestricted on economic grounds. These industries are generally commoditized or will likely become so in the near future. They typically feature gradual, incremental shifts in technology, pricing, and market share over time—not defensible moats or closing windows of opportunity for one country to gain enduring dominance. For example, the Biden administration should not add Honor, a smartphone maker spun off from Huawei, to the Entity List, as it is reportedly considering.22 Republican members of Congress have urged the designation, but their vague argument seems premised on the spurious notion that Honor operates “in a strategic sector.”23

AI. Artificial intelligence is frequently described as economically strategic—perhaps more often than any other technology area. Evangelists claim that AI will transform all other industries and become a primary determinant of competitive success. Even if this is true, competitive advantages in many aspects of AI do not appear to be very defensible. Foundational know-how proliferates widely (due to open, international academic ecosystems) and would be difficult to control (due to the relative ease of stealing or copying algorithms and training data), compared to more physically embodied trade secrets like semiconductor manufacturing equipment.24 Thus, U.S. government efforts to control AI research or software would often be ineffectual.

While China is quickly progressing in AI capabilities, there is also no clear sign that it verges on somehow dominating the industry. China does have certain advantages, including its ability to pool large stores of data with less concern for privacy and its access to massive, cheap sources of labor for data cleaning and preparation. Yet the United States still produces higher-quality AI research and has better access to data from key Western markets. At best, China seems capable of developing a modest lead in certain subdisciplines, like facial recognition, where Chinese advantages seem most relevant.25 But any such lead would be neither comprehensive (across all AI applications) nor permanent (foreclosing future U.S. competition).26

AI does appear to have some strategic terrain—bottlenecks in the AI value-chain where one nation might gain outsized advantages and seek to exclude its competitors. Semiconductors, discussed earlier, are one example. Another is the pool of high-end scientific and engineering talent. While the overall AI field is large, many of the most promising breakthroughs have come from a few individuals and companies, such as Alphabet’s DeepMind.27 Concentrating this talent together in one ecosystem seems to create disproportionate innovative benefits—although the effect is likely temporary, as AI innovations often proliferate widely within a few years. Thus, the United States should focus first and foremost on attracting the best AI talent and avoid decoupling the labor pool. This means ensuring that U.S. visa restrictions do not drive away top Chinese AI researchers from American universities and companies, unless a clear national security threat exists.

Key Offensive Policies

America’s ability to compete against China will depend much more on the health of the U.S. innovation ecosystem (so-called “offense”) than on any attempt to thwart or impede Chinese technological progress (“defense”). There are at least three major categories of offensive opportunities.28

First, Congress should greatly increase the amount of federal R&D spending. Such spending was historically pivotal in creating what we now know as Silicon Valley, but has atrophied in recent decades.29 The draft U.S. Innovation and Competition Act and America COMPETES Act, although not without problematic elements, mark important steps toward committing greater federal resources to R&D. Second, Congress should invest more in the social and physical infrastructure that supports technological innovation and access. Examples include STEM education (at all levels), STEM workforce training, a national research cloud, and rural broadband.30 Third, the Justice Department and the Federal Trade Commission should continue stepping up their antitrust scrutiny of the tech sector, and Congress should move forward with intelligent statutory reforms to promote competition. These would help ensure that the U.S. tech sector remains dynamic and innovative.


1 Remco Zwetsloot et al., “China is Fast Outpacing U.S. STEM PhD Growth,” Center for Security and Emerging Technology, August 2021,; “The Human Capital Report 2016,” World Economic Forum, 2016,; and “Gross domestic spending on R&D,” Organisation for Economic Co-operation and Development, 2021,

2 Robert Greene and Paul Triolo, “Will China Control the Global Internet Via its Digital Silk Road?,” SupChina, May 8, 2020,

3 For a review of this history, see James L. Schoff, “U.S.-Japan Technology Policy Coordination: Balancing Technonationalism With a Globalized World,” Carnegie Endowment for International Peace, June 29, 2020,

4 Mark Scott and Emily Birnbaum, “How Washington and Big Tech Won the Global Tax Fight,” Politico, June 30, 2021,

5 James Andrew Lewis, “Mapping the National Security Industrial Base: Policy Shaping Issues,” CSIS, May 19, 2021,

6 The blue-ribbon Cyberspace Solarium Commission expressed a version of this concern: “Chinese national companies like Huawei are part of an integrated strategy to use predatory pricing to dominate and eventually monopolize key information and communications technology supply chains. The goal is to drive non-Chinese alternatives out of business, leaving the Chinese Communist Party and its business allies with a stranglehold on the global supply chain.” “Final Report,” Cyberspace Solarium Commission, March 2020,

7 John S. McCain National Defense Authorization Act for Fiscal Year 2019, Public Law No. 115-232, § 1702(c)(1).

8 Commerce Department, “Review of Controls for Certain Emerging Technologies,” 83 Fed. Reg. 58,201 (November 19, 2018),

9 Peter Lichtenbaum, Victor Ban, and Lisa Ann Johnson, “Defining ‘Emerging Technologies’: Industry Weighs In on Potential New Export Controls,” China Business Review, April 17, 2019,; Paul H. DeLaney, “Business Roundtable Comments on the Advance Notice of Proposed Rulemaking (ANPRM) regarding the Review of Controls for Certain Emerging Technologies,” January 12, 2019,; and Linda Dempsey, “Comments of the National Association of Manufacturers on the Review of Controls for Certain Emerging Technologies (Docket BIS 2018-0024),” National Association of Manufacturers, January 9, 2019,

10 Council on Governmental Relations et al., “RIN 0694-AH80, Identification and Review of Controls for Certain Foundational Technologies ANPRM,” November 2020,

11 Steven W. Popper and Caroline Wagner, “Identifying Critical Technologies in the United States: A Review of the Federal Effort,” Journal of Forecasting 22, no. 2–3 (2003),; and Mary Ellen Mogee, Technology Policy and Critical Technologies: A Summary of Recent Reports (Washington DC: National Academies Press, 1991),

12 “National Critical Technologies Report—Appendix A: National Critical Technologies List,” White House, March 1995,

13 “National Strategy for Critical and Emerging Technologies,” White House, October 2020,

14 “Critical and Emerging Technologies List Update,” National Science and Technology Council, February 2022,

15 Emphasis in original.

16 “Asymmetric Competition: A Strategy for China & Technology,” China Strategy Group, Fall 2020,; and Jeffrey Ding and Allan Dafoe, “The Logic of Strategic Assets: From Oil to AI,” Security Studies 30, no. 2 (2021),

17 The United States does not have a single company that competes on a one-for-one basis with Huawei or ZTE in 5G telecommunications equipment markets. However, U.S. partners Finland (Nokia) and Sweden (Ericcson) do compete in aspects of these markets, and the United States has an interest in preventing Chinese dominance. Moreover, U.S. companies support various aspects of the 5G supply chain, and could do so more fully and effectively if the open, software-based 5G standard known as O-RAN continues to develop as a viable alternative to the closed, hardware-based systems sold by Huawei and ZTE.

18 Justin Hodiak and Scott W. Harold, “Can China Become the World Leader in Semiconductors?,” The Diplomat, September 25, 2020,; Christopher A. Thomas, “Lagging but Motivated: The State of China’s Semiconductor Industry,” Brookings TechStream (blog), Brookings Institution, January 7, 2021,; Arjun Kharpal, “China Is Ramping Up Its Own Chip Industry Amid a Brewing Tech War,” CNBC, June 4 2019,; and Jordan Schneider, “China’s Chip Industry: Running Faster But Still Falling Behind,” Rhodium Group, April 22, 2021,

19 Alexandra Alper, Toby Sterling, and Stephen Nellis, “Trump Administration Pressed Dutch Hard to Cancel China Chip-Equipment Sale: Sources,” Reuters, January 6, 2020,; and Justin Hodiak and Scott W. Harold, “Can China Become the World Leader in Semiconductors?,” The Diplomat, September 25, 2020,

20 Karen Freifeld, “Huawei Gets U.S. Approvals to Buy Auto Chips, Sparking Blow Back,” Reuters, August 25, 2021,

21 For an explanation of this issue, see Evan Burke, “Trump-Era Policies Toward Chinese STEM Talent: A Need for Better Balance,” Carnegie Endowment for International Peace, March 25, 2021,

22 Ellen Nakashima and Jeanne Whalen, “Key Security Agencies Split Over Whether to Blacklist Former Huawei Smartphone Unit,” Washington Post, September 19, 2021,

23 Michael McCaul et al., letter to Gina Raimondo, August 6, 2021,

24 Christine Fox, “An Entwined AI Future: Resistance Is Futile,” 2020, JHU APL,

25 Regarding China’s potential advantages in certain AI sub-fields, one recent study noted that “a disproportionate share of China’s highly cited and top-venue [AI] publications include publications on general-purpose computer vision research, as well as applications of AI to surveillance and industry.” Ashwin Acharya and Brian Dunn, “Comparing U.S. and Chinese Contributions to High-Impact AI Research,” Center for Security and Emerging Technology, January 2022,

26 Rand Waltzman et al., “Maintaining the Competitive Advantage in Artificial Intelligence and Machine Learning,” RAND Corporation, 2020,

27 Sam Shead, “Why the Buzz Around DeepMind Is Dissipating as It Transitions From Games to Science,” CNBC, June 5, 2020,

28 A much more detailed and comprehensive set of proposals, related primarily but not exclusively to AI, can be found in the report of the National Security Commission on Artificial Intelligence. “Final Report,” National Security Commission on Artificial Intelligence, March 2021,

29 Robert D. Atkinson, “The Case for a National Industrial Strategy to Counter China’s Technological Rise,” Infornation Technology & Innovation Foundation, April 13, 2020,

30 Robert D. Atkinson, “The Case for a National Industrial Strategy to Counter China’s Technological Rise,” Infornation Technology & Innovation Foundation, April 13, 2020,