Table of Contents

Risks of Interdependence

U.S. officials have become increasingly vocal in warning of Chinese government efforts to influence American politics and society.1 Several trends underlie this concern. First, Russia’s interference in the 2016 U.S. presidential election brought much greater attention to the overall threat of foreign influence. Subsequent events, particularly the COVID-19 “infodemic” and the U.S. Capitol insurrection, further highlighted the fragility of America’s political-informational ecosystem and its susceptibility to damaging manipulation.

At the same time, a more assertive China has increasingly sought to shape political narratives beyond its borders, especially on China-related issues. Much of this activity is overt—including Beijing’s “wolf warrior diplomacy,” its nationalistic state-sponsored media, and its punishment of foreign companies whose speech offends the Chinese Communist Party. But some efforts are covert—ranging from traditional influence (for example, cultivating agents within foreign political circles) to modern digital techniques (for example, fabricating armies of fake social media accounts that harass and vilify dissidents).2

China’s foreign influence efforts have often focused closer to home, on targets such as Taiwan and Australia.3 Nevertheless, in 2021 the U.S. Intelligence Community assessed that “Beijing has been intensifying efforts to shape the political environment in the United States to promote its policy preferences, mold public discourse, pressure political figures whom Beijing believes oppose its interests, and muffle criticism of China on such issues as religious freedom and the suppression of democracy in Hong Kong.”4 According to the IC, China “considered but did not deploy influence efforts intended to change the outcome of the [2020] US presidential election.”5 Beijing apparently judged that the risks outweighed the benefits. This calculus may well change in the future, particularly if U.S.-China relations continue to deteriorate.

U.S. policymakers worry that Beijing could pressure the makers of popular Chinese apps like TikTok or WeChat to support covert influence campaigns that target Americans.

U.S. policymakers worry that Beijing could pressure the makers of popular Chinese apps like TikTok or WeChat to support covert influence campaigns that target Americans. Such a campaign might involve artificially promoting and/or suppressing certain content, perhaps leveraging the apps’ capability to microtarget specific audiences. WeChat—which is popular among the global Chinese diaspora—already censors topics such as the Tiananmen Square massacre and the Falun Gong religious movement.6 TikTok has acknowledged doing the same, but claims that it stopped in 2019.7

Risks and Limitations of Defensive Measures

In theory, these concerns could justify major U.S. government restrictive measures such as app bans. In practice, however, there are good reasons for U.S. policymakers to think twice. To begin with, social media–based influence operations by China and other foreign governments may not actually be very effective. Researchers have struggled to find strong evidence that such operations can measurably alter their targets’ beliefs and actions over time. A meta-analysis by Princeton University found only one high-quality empirical study on the question.8 That study examined efforts by Russia’s Internet Research Agency during the 2016 U.S. presidential election and found no effect on American Twitter users’ political beliefs.

Although future research may identify stronger causal effects, it is worth keeping in mind the difficulties of large-scale public persuasion. To swing a U.S. presidential election, for example, Chinese influence actors would need to sway the small number of persuadable voters, or alter the turnout of voters, in battleground states. But Beijing would face stiff competition along the way. A Chinese influence campaign would be operating amid a cacophony of other voices—including political candidates and parties, community leaders, activists, traditional media commentary, and authentic citizen views—that dominate online as well as TV, radio, print, and word-of-mouth discourse. Domestic voices tend to have far more resources (billions of dollars are spent during a presidential election cycle), greater political sophistication, and thicker networks than even the most well-crafted foreign personas.

Just as domestic actors are the predominant voices in American politics, domestic players are also the main sources and amplifiers of political disinformation. Before TikTok entered the American market, U.S.-based platforms like Facebook, Twitter, and YouTube already provided fertile ground for false, polarizing, and destructive political discourse. While these platforms have taken many actions to address influence operations, the problem seems to be growing even faster.9 Fundamentally, user-generated content is produced on a scale that overwhelms existing content moderation tools, and platforms lack the financial and political incentives to undertake wholesale product redesigns to reduce the spread of harmful content. Market power is one part of the problem. The size of some major platforms helps to insulate them from pressure by users, advertisers, political leaders, activists, and employees to take stronger action against influence operations and other damaging content.

In this context, competitive pressure from Chinese apps like TikTok may have beneficial effects. TikTok is the most significant competitive threat to emerge in the American social media landscape in years. As such, its presence might help spur U.S.-based platforms to take stronger action against disinformation and other influence operations to burnish their reputations among advertisers, users, and outside stakeholders. Indeed, American activists and NGOs concerned about harmful online content have begun to explicitly compare TikTok’s efforts against those of U.S.-based platforms.10 This suggests that TikTok’s presence has helped to intensify a reputational contest among platforms that could, if combined with regulatory and other pressure, raise the bar for responsible policies and practices by all players.

Slow-rolling efforts to shape Americans’ general views of China and China-related policy are more readily detected and countered without resort to government tech controls.

A sound U.S. policy on Chinese influence operations would place companies like TikTok in the context of the larger American political-informational ecosystem. Seen in that light, Chinese tech companies play a limited and, perhaps, not entirely harmful role. Restrictive measures to counter Chinese influence operations should therefore be carefully vetted and focus on the highest-impact, most plausible threats.

Recommended Policies and Processes

Washington has a variety of tools to combat Chinese influence operations, but it should reserve technology restrictions such as app bans for the most serious risks. These would include the risk that China successfully alters a national election outcome or greatly reduces public confidence in an election. Potential influence operations with life-and-death consequences, such as those that markedly increase vaccine hesitancy during a pandemic, would also justify strong controls. However, long-term influence operations on less sensitive topics can often be managed in other ways. The bulk of Chinese influence activity in the United States seems aimed at shaping Americans’ general views of China and China-related policy.11 These slow-rolling persuasion campaigns, while troubling, are no emergency. They are more readily detected and countered without resort to government controls.

It is unclear whether China currently has the ability to achieve any of the most dangerous influence outcomes, such as swinging an election. The U.S. government should conduct a careful, fact-based assessment to guide its use of technology controls. The Intelligence Community can help by estimating the Chinese government’s capability and willingness to subvert Chinese commercial technology to influence Americans. But a China-focused intelligence assessment is only part of what policymakers would need. U.S. policymakers must also understand the American political and societal factors that would determine whether Chinese influence operations ultimately succeed or fail. This analysis would be crucial to properly size up the threat and weigh policy responses, yet the Intelligence Community and other government agencies lack the authority and expertise to conduct such an assessment.

To supplement the IC’s analysis, the president could convene an outside advisory group of political scientists, communications experts, influence operations researchers, and technologists, perhaps under the aegis of the National Academies of Sciences, Engineering, and Medicine. This panel would examine the U.S. domestic environment’s susceptibility to Chinese digital influence operations. For example, it might consider whether and how such operations could effectively persuade key voting constituencies or influence their turnout, considering factors like the partisan balance in swing states and the responsiveness of various constituencies to targeted digital campaigns. A government-sponsored analysis of this kind would need to be carefully designed to prevent real and perceived impingements on Americans’ civil liberties.12

Case Studies

TikTok. The most significant U.S. restrictions aimed at thwarting Chinese influence operations were Trump’s executive orders attempting to ban TikTok and force its sale. These orders—which have never been implemented—were not justified based on publicly available evidence about influence threats. While Trump vaguely claimed that TikTok could “be used for disinformation campaigns that benefit the Chinese Communist Party,” his administration offered no analysis of how effective these campaigns might be.13

Biden rescinded the TikTok ban. He replaced it with a new mechanism, the Commerce Department’s ICTS supply chain security rule, to evaluate any Chinese software and hardware popularly used in the United States. Later, he published “a criteria-based decision framework” to help guide the Commerce Department’s review of so-called “connected software applications” such as TikTok.14 However, officials have not publicly confirmed whether TikTok is currently being investigated under the ICTS process.15

The Department of Commerce should carry out such a review, if it has not already started one. Biden’s criteria offer a helpful starting point but should be refined to more specifically assess TikTok’s threat as a medium for Chinese influence operations. Outside experts should draw on the best data and science to answer key questions including: whether TikTok’s user base contains a large number of swing state voters; whether political content on TikTok content appears highly influential with a critical mass of those voters; whether corporate firewalls cannot reliably prevent Beijing from hijacking the platform in an undetected way during the course of an election; and whether the threat of Chinese influence operations via TikTok outweighs any benefits that TikTok may have on U.S. political discourse, including from competitive pressures on American tech platforms.

Long-term influence. Beyond TikTok, many proposals to limit Chinese influence capabilities in the United States do not focus on high-consequence, time-critical processes like elections. Instead, there is often worry that Beijing may gradually sway Americans’ views about China-related policies. Confucius Institutes (Chinese public diplomacy initiatives embedded in U.S. universities) are frequent bogeymen, as is Chinese influence over U.S. entertainment sectors, like filmmaking and sports. CFIUS is reportedly in talks with Chinese tech giant Tencent about its ownership stakes in major U.S. video game developers.16 At some point, relationships between Chinese tech companies and U.S. streaming platforms—like Netflix-Baidu and HBO-Tencent—will likely come under scrutiny. But none of these arrangements seem to represent the kind of urgent influence threat that justifies forceful U.S. government controls.

Key Offensive Policies

While the U.S. government should continue to monitor and disrupt Chinese influence activities, its top priority must be restoring health to America’s domestic information ecosystem. Washington must recognize that disinformation flourishes due to deep-seated and largely homegrown trends—in American politics, society, economy, and law—that have co-evolved over decades and become mutually reinforcing. Key factors include the TV and online media landscape (segmenting Americans into ideological echo chambers), social media business models (maximizing user engagement and enabling microtargeting), and political party dynamics (such as geographic sorting, gerrymandering, and primary election rules). Large-scale progress in combating disinformation would require profound national reforms in these and other arenas. The goal would be to disincentivize the production, amplification, and consumption of disinformation from all sources—not just China.

The top priority must be restoring health to America’s domestic information ecosystem. This would disincentivize the production, amplification, and consumption of disinformation from all sources—not just China.

True reform would be an extremely daunting task. The federal government’s role in combating disinformation is poorly defined and heavily constrained by laws, norms, and political obstacles. Its tools are often tactical in nature (like sanctions) and oriented toward foreign threats (as with the Foreign Agents Registration Act). Federal overreach could actually worsen political distrust or create harmful precedents that future administrations could abuse. In fact, some of the most dangerous disinformers have been federal officeholders and candidates.

That said, experts have proposed a raft of policy ideas that the U.S. government could either implement or help to coordinate. These include strengthening regulation of online platforms; reforming campaign finance, election advertising, and redistricting laws; funding media literacy education; creating new public-private grant programs for journalists; and funding and facilitating basic research on influence operations.17 Such policies have not been rigorously tested. In fact, there is very little empirical evidence about the impacts of influence operations or the effectiveness of countermeasures.18 Still, improving the U.S. domestic information environment would be much more effective in curbing Beijing’s influence operations than any China-specific measures. Moreover, these policies would help to address domestic disinformation, a far more serious problem.

Notes

1 Christopher Wray, “The Threat Posed by the Chinese Government and the Chinese Communist Party to the Economic and National Security of the United States” (video lecture, Hudson Institute, Washington, DC, July 7, 2020), https://www.fbi.gov/news/speeches/the-threat-posed-by-the-chinese-government-and-the-chinese-communist-party-to-the-economic-and-national-security-of-the-united-states.

2 Bethany Allen-Ebrahimian and Zach Dorfman, “Suspected Chinese Spy Targeted California Politicians,” Axios, December 8, 2020, https://www.axios.com/china-spy-california-politicians-9d2dfb99-f839-4e00-8bd8-59dec0daf589.html; Bethany Allen-Ebrahimian and Zach Dorfman, “Republican Donor Cindy Yang Linked to Chinese Influence Machine,” Foreign Policy, March 12, 2019, https://foreignpolicy.com/2019/03/12/962067-china-unitedfront-corruption-scandal/; and Bethany Allen-Ebrahimian, “China Built an Army of Influence Agents in the U.S.,” Daily Beast, July 18, 2018, https://www.thedailybeast.com/how-china-built-an-army-of-influence-agents-in-the-us.

3 Joshua Kurlantzick, “How China Is Interfering in Taiwan’s Election,” Council on Foreign Relations, November 7, 2019, https://www.cfr.org/in-brief/how-china-interfering-taiwans-election; Amy Searight, “Countering China’s Influence Operations: Lessons From Australia,” Center for Strategic and International Studies, May 8, 2020, https://www.csis.org/analysis/countering-chinas-influence-operations-lessons-australia; and Alex Joske, Lin Li, Alexandra Pascoem and Nathan Attrill, “The influence environment: A Survey of Chinese-language Media in Australia,” Australian Strategic Policy Institute, December 2020, https://s3-ap-southeast-2.amazonaws.com/ad-aspi/2020-12/The%20influence%20environment.pdf?Mjytg1N_V2azySJlJoahgWzMOqxEpMYk.

4 “Annual Threat Assessment of the US Intelligence Community,” Director of National Intelligence, April 9, 2021, https://www.dni.gov/files/ODNI/documents/assessments/ATA-2021-Unclassified-Report.pdf.

5 “Intelligence Community Assessment: Foreign Threats to the 2020 US Federal Elections,” National Intelligence Council, March 10, 2021, https://www.dni.gov/files/ODNI/documents/assessments/ICA-declass-16MAR21.pdf.

6 Alex Joske, Lin Li, Alexandra Pascoem, and Nathan Attrill, “The Influence Environment: A Survey of Chinese-language Media in Australia,” Australian Strategic Policy Institute, December 2020, https://s3-ap-southeast-2.amazonaws.com/ad-aspi/2020-12/The%20influence%20environment.pdf?Mjytg1N_V2azySJlJoahgWzMOqxEpMYk.

7 Alex Hern, “Revealed: How TikTok Censors Videos That Do Not Please Beijing,” Guardian, September 25, 2019, https://www.theguardian.com/technology/2019/sep/25/revealed-how-tiktok-censors-videos-that-do-not-please-beijing; Isobel Asher Hamilton, “A senior TikTok executive admitted the company used to censor content critical of China, ‘specifically with regard to the Uighur situation,’” Business Insider, November 5, 2020, https://www.businessinsider.com/tiktok-censor-china-critical-content-uighur-uighurs-2020-11.

8 Jon Bateman, Elonnai Hickock, Laura Courchesne, Isra Thange, and Jacob N. Shapiro, “Measuring the Effects of Influence Operations: Key Findings and Gaps From Empirical Research,” Carnegie Endowment for International Peace, June 28, 2021, https://carnegieendowment.org/2021/06/28/measuring-effects-of-influence-operations-key-findings-and-gaps-from-empirical-research-pub-84824.

9 Major platform efforts include revisions to community standards, enforcement actions taken against proscribed activity, and adjustments to the functionality of platform products. For overviews of each of these three areas, see Jon Bateman, Natalie Thompson, and Victoria Smith, “How Social Media Platforms’ Community Standards Address Influence Operations,” Carnegie Endowment for International Peace, April 1, 2021, https://carnegieendowment.org/2021/04/01/how-social-media-platforms-community-standards-address-influence-operations-pub-84201; “Disinfodex,” Berkman Klein Center at Harvard University and the Ethics and Governance of Artificial Intelligence Fund at The Miami Foundation, 2020, https://disinfodex.org/; and Kamya Yadav, “Platform Interventions: How Social Media Counters Influence Operations,” Carnegie Endowment for International Peace, January 25, 2021, https://carnegieendowment.org/2021/01/25/platform-interventions-how-social-media-counters-influence-operations-pub-83698.

10 “One Year After Stop Hate for Profit: Platforms’ Progress,” Stop Hate for Profit, June 16, 2021, https://www.stophateforprofit.org/platforms-progress-year-later; and Jessica Guynn, “Facebook Winning War on COVID-19 Vaccine Lies, Hoaxes and Conspiracies. Twitter and TikTok? Not So Much, Report Says,” USA Today, May 7, 2021, https://www.usatoday.com/story/tech/2021/05/07/facebook-twitter-tiktok-covid-vaccine-conspiracy-theories-lies-hoaxes/4994731001/.

11 Olivia Solon and Ken Dilanian, “China’s Influence Operations Offer a Glimpse Into the Future of Information Warfare,” NBC News, October 21, 2020, https://www.nbcnews.com/business/business-news/china-s-influence-operations-offer-glimpse-future-information-warfare-n1244065.

12 Betsy Woodruff Swan, “DHS Stands Up Domestic Terror Intelligence Team,” Politico, May 11, 2021, https://www.politico.com/news/2021/05/11/dhs-domestic-terror-intelligence-487145.

13 Executive Order 13942, “Addressing the Threat Posed by TikTok, and Taking Additional Steps To Address the National Emergency With Respect to the Information and Communications Technology and Services Supply Chain,” 85 Fed. Reg. 48,637 (August 6, 2020), https://www.federalregister.gov/documents/2020/08/11/2020-17699/addressing-the-threat-posed-by-tiktok-and-taking-additional-steps-to-address-the-national-emergency.

14 “Fact Sheet: Executive Order Protecting Americans’ Sensitive Data from Foreign Adversaries,” White House, June 9, 2021, https://www.whitehouse.gov/briefing-room/statements-releases/2021/06/09/fact-sheet-executive-order-protecting-americans-sensitive-data-from-foreign-adversaries/; and Executive Order 14034, “Protecting Americans’ Sensitive Data From Foreign Adversaries,” 86 Fed. Reg. 31,423 (June 9, 2021), https://www.federalregister.gov/documents/2021/06/11/2021-12506/protecting-americans-sensitive-data-from-foreign-adversaries.

15 Marco Rubio, “Rubio Requests Biden Administration’s Plan for TikTok, Protecting Americans’ Data,” press release, October 12, 2021, https://www.rubio.senate.gov/public/index.cfm/2021/10/rubio-requests-biden-administration-s-plan-for-tiktok-protecting-americans-data.

16 Greg Roumeliotis and Echo Wang, “China’s Tencent in Talks with U.S. to Keep Gaming Investments -Sources,” Reuters, May 5, 2021, https://www.reuters.com/technology/exclusive-chinas-tencent-talks-with-us-keep-gaming-investments-sources-2021-05-05/.

17 Vishnu Kannan, forthcoming paper, Carnegie Endowment for International Peace; and Kamya Yadav, “Countering Influence Operations: A Review of Policy Proposals Since 2016,” Carnegie Endowment for International Peace, November 30, 2020, https://carnegieendowment.org/2020/11/30/countering-influence-operations-review-of-policy-proposals-since-2016-pub-83333.

18 Jon Bateman, Elonnai Hickock, Laura Courchesne, Isra Thange, and Jacob N. Shapiro, “Measuring the Effects of Influence Operations: Key Findings and Gaps From Empirical Research,” Carnegie Endowment for International Peace, June 28, 2021, https://carnegieendowment.org/2021/06/28/measuring-effects-of-influence-operations-key-findings-and-gaps-from-empirical-research-pub-84824; and Jon Bateman, Elonnai Hickock, Jacob N. Shapiro, Laura Courchesne, and Julia Ilhardt, “Measuring the Efficacy of Influence Operations Countermeasures: Key Findings and Gaps From Empirical Research,” Carnegie Endowment for International Peace, September 21, 2021, https://carnegieendowment.org/2021/09/21/measuring-efficacy-of-influence-operations-countermeasures-key-findings-and-gaps-from-empirical-research-pub-85389.