Hoover Institution (Stanford, CA) – The Hoover Institution’s project on China’s Global Sharp Power, chaired by Senior Fellow Larry Diamond and managed by Research Fellow Glenn Tiffert, cohosted a virtual conference about how the People’s Republic of China’s advancements in artificial intelligence (AI) are used to repress its population and support authoritarian modes of governance across the world.
The four-day program, The Rise of Digital Authoritarianism: China, AI, and Human Rights, was hosted in partnership with the Digital Policy Incubator at the Freeman Spogli Institute, Stanford’s Center for Human-Centered Artificial Intelligence (HAI), and the Human Rights Foundation.
Keynote remarks were given by Condoleezza Rice, Hoover Institution director; Eric Schmidt, former Google chairman and CEO; Michael Brown, director of the Defense Innovation Unit at the US Department of Defense; Audrey Tang, Taiwan’s digital minister; and Fei-Fei Li, co-director of HAI.
Condoleezza Rice Says AI Allows China to “Dream Big”
In opening remarks at the conference on September 29, Condoleezza Rice argued that for authoritarian regimes like the Chinese Communist Party, the unrestrained application of AI and other emerging technologies has allowed them to “dream big.” She maintained that Beijing has not only created an Orwellian state apparatus that has imposed mass surveillance on its citizens and ethnic minorities, but has also enabled the People’s Liberation Army to strengthen its forces, as evident in its assertiveness in the South China Sea and the Strait of Taiwan.
Rice cautioned against US leaders embracing technological innovation in order to capably compete with Beijing and abandoning institutional constraints that account for ethics, individual liberty, and concerns over personal privacy.
“We need to have a concerted effort on behalf of free peoples to make sure that the digital authoritarians don’t win,” Rice said. “They can’t win the race for this technology, because whoever wins this race is going to have a leg up on shaping the international system.”
China’s Use of Technology Is Not Just Authoritarian, It Is Totalitarian
Rice’s remarks were followed by a panel discussion featuring Bethany Allen-Ebrahimian, China reporter for Axios; Paul Mozur, Asia technology correspondent for the New York Times; Glenn Tiffert, Hoover Institution research fellow; and Xiao Qiang, editor-in-chief of the China Digital Times.
The participants explained how China deploys AI, machine learning, and facial recognition technology for large-scale analysis of data for mass surveillance and the implementation of a national social credit score system. They noted that the data to support these systems is aggregated from security cameras, internet and mobile phone activity, financial transactions, informants, and other sources.
The panel also detailed the Chinese government’s Integrated Joint Operations Platform (IJOP), a software application that centralizes data about individual behavior and has been used to aid police in the arrest, repression, and incarceration of segments of the Uighur Muslim population in the Xinjiang autonomous region. They concluded that while IJOP and other methods of surveillance are imperfect and frequently crude in their implementation, they fulfill the party’s aspirations for totalitarian rule by coercing ethnic minorities and other citizens into compliance.
A Rivalry Partnership?
On the second day of the virtual conference, October 1, Eric Schmidt told Eileen Donahoe, director of the Digital Policy Incubator, that US leaders should use a “scalpel” instead of a “sledgehammer” when developing and executing policies aimed at preventing China from overtaking US technological supremacy.
Schmidt explained that while the US is superior to China in AI technology, this lead will diminish within two years. He cautioned against drastic decoupling of supply chains, export controls on semiconductors (an industry in which the US remains dominant), and excessive restrictions against US-Chinese research collaborations and Chinese students who want to study at American universities. He maintained, however, that US policy makers should remain vigilant against Chinese espionage activities and effectively enforce rules and regulations.
Schmidt added that American innovation should be imbued with democratic values in cooperation with Chinese counterparts—a strategy he called a “rivalry partnership.” This collaboration, Schmidt believes, would enable the United States to mitigate China’s efforts in replicating American technology as well as curtail Beijing’s ambitions to expand its global market share in 5G telecommunications platforms.
Risking a Splinternet
The second day of programming also included a discussion about the ethics of doing business in China featuring Mary Hui, Hong Kong–based technology and business reporter of Quartz; Megha Rajagopolan, international correspondent of BuzzFeed News; and Alex Stamos, director of Stanford’s Internet Observatory.
The three explained how China has become globally dominant in software industries and a central node in key supply chains that are extremely difficult to shift to other countries or dismantle entirely without causing major global economic disruption.
The participants also raised concerns over a splintering of global internet governance, especially in light of the Trump administration’s recent attempts to ban the Chinese social media app TikTok. They explained that such use of presidential power may lead China to retaliate by expropriating US companies. The panel instead advocated for a US-led mode of internet governance that enforces human rights standards and secures privacy protections for users.
Increasing Government Supported R&D Is Critical for US Security
The third day of the conference, October 6, featured a keynote discussion by Michael Brown, director of the Defense Innovation Unit at the US Department of Defense.
Brown argued that in order for the United States to maintain its technological competitiveness with China, the federal government must increase funding for research and development in emerging technologies from the current 0.7 percent to 2 percent of GDP, a level not reached since the 1970s. Brown explained that government support is critical, because it provides industry with the proper incentives to focus on long-term national priorities reminiscent of the Manhattan Project, the space program, and the production of advanced weaponry during the Cold War rivalry with the Soviet Union.
Brown also maintained that India is an important strategic ally for the United States, contending that the South Asian country’s democratic principles, its younger demographics, and the value it places on education sets it on a path to surpass China technologically in less than a decade.
A Competition of Value Systems, Not Superpowers
Brown’s presentation was followed by a conversation about the geopolitical implications of China’s emergence as an exporter of AI and mass-surveillance technologies. Discussants included Steven Feldstein, senior fellow at the Carnegie Endowment for International Peace; Lindsey Gorman, fellow for emerging technologies at the German Marshall Fund; and Maya Wang, senior researcher at Human Rights Watch.
The panel agreed that it would be advantageous for American policy makers to respond to this challenge not as a contest between the United States and China but rather as a competition between democratic and authoritarian systems. The participants held that the latter approach would enable the United States to more persuasively attract the cooperation of trade and security partners whose shared devotion to core values of human rights and freedom could surmount the economic benefits of closer ties to Beijing.
Taiwan’s Response China’s Emergence as a Tech Leader
The final day of the conference, October 9, was opened by Audrey Tang, the digital minister of Taiwan. Tang described the actions Taipei is taking to defend the security of its telecommunications networks from Beijing’s aggression. She explained that Beijing’s mode of technological governance is a “rule by law” standard based on the dictates, biases, and interests of the Communist Party. She warned that Beijing was attempting to reshape international norms and subvert transparency and the rule of law in democracies across the world by exporting technologies that are capable of being used for mass surveillance and human rights abuses.
She also explained that Taiwan was helping to lead a cooperative effort with several other countries to ensure that supply chains of hardware and software are untainted by Beijing’s influence and that data privacy is safeguarded against encroachments by Chinese state-controlled technology providers.
The Democratic Response to Digital Authoritarianism
The final panel discussion of the conference covered what democracies can do to curb the expansion of authoritarianism through the application of AI and other emerging technologies.
Participants included Christopher Balding, associate professor of Fulbright University, Vietnam; Anja Manuel, cofounder of Rice, Hadley, Gates, and Manuel; and Chris Meserole, deputy director of the Artificial Intelligence and Emerging Technologies Initiative at the Brookings Institution.
The panelists held that democratic countries should affirm their values by working together on the development of clear and transparent governance over the use of technology; collectively supporting open research while enforcing against such crimes as intellectual property theft; and maintaining an edge over China in the manufacturing of semiconductors.
A Multistakeholder Approach to AI
Fei-Fei Li, the codirector of Stanford HAI concluded the conference by stressing ethical considerations in the development of AI technology.
She explained that while AI has many positive applications in industries such as health care, education, and manufacturing, it can also be used for nefarious ends by authoritarian rulers including the promotion of self-serving political narratives and the amplification of racial and cultural biases.
Li concluded that design of machines is an extension of the human condition. If a society respects human rights, values diversity in its population, and is committed to protecting the freedoms of all its citizens, then those principles will be embedded in the design and application of machines.
“In order to build a future in which AI supports democracy, the democracies of today must support AI,” Li said.
Click here to watch the entire conference.
