Citizen Lab submission to the Congressional-Executive Commission on China about the State of Human Rights in China

Emile Dirks, Research Associate at the Citizen Lab, prepared a written submission for the Congressional-Executive Commission on China (CECC) about the state of human rights in the country. The CECC was established by Congress in October 2000, with the legislative mandate to monitor human rights and the development of the rule of law in China, and to present an annual report to the President and the Congress. The Citizen Lab sent in this submission on February 1, 2024. 

Emile Dirks’ submission focuses on Chinese government-backed online censorship and its impact on the freedom of expression and opinion of people accessing the internet from within China or using China accessible online platforms. Dirks argues that not only the Chinese government, but also US-based firms, are complicit in the political and religious censoring of content on China-accessible platforms. In addition to online risks, the submission emphasizes how online censorship impedes personal freedom and can be connected to transnational repression. 

Finally, the paper makes recommendations for what the US government may do to protect persons affected by online censorship both within and outside of China. This includes holding Chinese and US firms accountable for their involvement in online censorship and assisting victims of digital abuse and intimidation. 

The Citizen Lab’s full submission to the Congressional Executive Commission on China is available on the website of CECC and reproduced below. On February 1, Emile Dirks participated remotely from Toronto in a hearing on “The PRC’s Universal Periodic Review and the Real State of Human Rights in China” hosted by the CECC. You can watch the video here. 

Hearing Before the Congressional-Executive Commission on China

“The PRC’s Universal Periodic Review and the Real State of Human Rights in China”

Thursday, February 1, 2024

Written Testimony by Dr. Emile Dirks

Research Associate at The Citizen Lab at the University of Toronto

Representative Smith, Senator Merkley, and distinguished Members of the Commission, thank you for holding this important hearing on the state of human rights in the People’s Republic of China and for the opportunity to testify. The conclusion of the United Nations Human Rights Council’s Universal Periodic Review (UPR) Working Group’s review of China provides an excellent opportunity to reflect on the current state of human rights in China.

My testimony today draws upon the work of myself and other researchers at the Citizen Lab. The Citizen Lab is an interdisciplinary research laboratory based at the Munk School of Global Affairs & Public Policy at the University of Toronto, focused on research, development, and strategic policy and legal engagement at the intersection of information and communication technologies, human rights, and global security.¹ Citizen Lab research has explored transnational repression, spyware, censorship, algorithmic policing, and biometric surveillance.²

Today I will focus my remarks on a particular aspect of China’s human rights record: Chinese state-backed online censorship. My testimony will highlight three key points concerning online censorship. One, state-backed online censorship profoundly impacts Chinese citizens’ freedom of opinion and expression, as well as the freedom of opinion and expression of those accessing the internet from within China or using China-accessible online platforms. Two, both Chinese and US companies contribute to online censorship on China-accessible platforms. And three, online censorship is linked to repression inside China and transnational repression outside China, both of Chinese citizens and Chinese, Hong Kong, Uyghur, Tibetan, and other diaspora members.

Drawing on these three points, I will conclude with three recommendations for how the United States government can demand accountability from perpetrators and provide assistance to victims. One, the United States government should publicly request that Microsoft, Apple, and other US companies explain how they implement political and religious censorship on their platforms in China. Two, the United States government should publicly request that Microsoft explain how political and religious censorship was applied to the search suggestions of users outside China and what safeguards will ensure this will not reoccur. And three, the US government should provide training to relevant US government officials, including law enforcement and immigration authorities, to recognize digital transnational repression and properly assist victims and their families.

Part One: State-backed Online Censorship

The Chinese government severely restricts Chinese citizens’ freedom of opinion and expression through online censorship, as detailed by the United Nations Committee on Economic, Social and Cultural Rights and stakeholders’ submissions for China’s most recent periodic review.³ Using a sophisticated filtering system known as the “Great Firewall,” authorities block access to thousands of websites which provide information which challenges the preferred narratives of the Chinese government.⁴

One of the clearest measurements of state-mandated censorship comes from Great Firewall Watch, a platform created by researchers at Stony Brook University, the University of Massachusetts – Amherst, the University of California, Berkeley, and the Citizen Lab at the University of Toronto.⁵ Since its inception in March 2020, GFWatch.org has discovered more than 640,000 blocked domains.⁶ GFWatch.org can also be used to test whether a particular domain is accessible within China.⁷ Blocked domains include the website of the Congressional-Executive Commission on China, as well as the websites of groups whose members have previously testified before the Commission, including Tibet Action Institute, Human Rights Watch, Hong Kong Democracy Council, Uyghur Human Rights Project, and the Australian Strategic Policy Institute.

Blocking websites is not the only way that the Chinese government attempts to restrict freedom of opinion and expression. Online censorship is pervasive even on platforms accessible in China. Numerous Chinese government offices participate in online censorship, including the Cyberspace Administration of China and the Ministry of Public Security. To clarify what broad categories of online material are prohibited, the Chinese government has issued a number of documents, including the Measures for the Administration of Security Protection of Computer Information Networks with International Interconnections (1997), the Cybersecurity Law (2017), Norms for the Administration of Online Short Video Platforms and Detailed Implementation Rules for Online Short Video Content Review Standards (2019), and Provisions on the Governance of the Online Information Content Ecosystem (2020).⁸ Prohibited content listed in these documents includes “content harming the image of revolutionary leaders or heroes and martyrs” and information which is “damaging the reputation or interests of the state” or “detrimental to state religious policies, propagating heretical or superstitious ideas.”⁹ The Chinese government also routinely conducts “internet purification campaigns” by which state organs compel websites, platforms, and accounts to remove prohibited content and punish violators through warnings or administrative or criminal penalties.¹⁰

Yet while government authorities stipulate what broad categories of content is prohibited, it is technology companies which are responsible for day-to-day censorship. Technology companies operating in China are required to ensure that content which appears on their platforms complies with legal requirements or political directives from the Chinese state. Companies which fail to moderate content on their platforms can be fined or have their business licenses revoked.¹¹ This form of intermediary liability or corporate “self-discipline” is a characteristic feature of information control and online censorship in China.¹²

Citizen Lab researchers have discovered over 60,000 censorship rules on eight China-accessible search platforms: Baidu, Baidu Zhidao, Bilibili, Microsoft Bing, Douyin, Jingdong, Sogou, and Weibo.¹³ Examples of censored content covered by these rules include various creative homoglyphs for the name “Xi Jinping,” references to the June 4 massacre, material related to religious communities, and criticisms of the Communist Party. Citizen Lab research also demonstrates that platforms institute different levels of censorship which fully or partially censor search results for key terms. Partial or “soft” censorship provides results from authorized sources like Chinese government websites or state media, while full or “hard” censorship provides no results.

Citizen Lab researchers have also detailed how China-accessible platforms including WeChat censor discussion of political events. These events include activism in Hong Kong, crackdowns on human rights lawyers, the COVID-19 pandemic, and the deaths of Nobel Peace Prize winner Liu Xiaobo and former premier Li Keqiang.¹⁴ Results from these research investigations demonstrate how China-accessible platforms suppress politically sensitive information and promote narratives favourable to the Chinese state.

Part Two: The Role of US-based Companies in Online Censorship

While Chinese tech companies are the key players in online censorship in China, US companies are also involved. For instance, in 2018 leaked documents revealed that Google was planning to release an app in China that would implement political censorship, a plan they abandoned in 2019 after criticism from within and outside the company.¹⁵ On their China-accessible platforms, US companies have imposed restrictions on political and religious content. And like the restrictions imposed by Chinese counterparts, those imposed by US companies have impacted users both inside China and in other world regions, including the United States.

Citizen Lab research shows that the Chinese version of Microsoft’s Bing, the only major non-Chinese search engine accessible in China, engages in extensive censorship.¹⁶ In China, Bing only displays results for censored search queries from authorized websites, such as government and state media websites. Like Chinese search platforms, Bing’s censorship rules target political material related to Xi Jinping, religious material, references to Nobel Peace Prize winner Liu Xiaobo, and terms related to the June 4 massacre. Compared with Baidu, Bing’s political censorship rules are also broader, affect more search results, and lead to search results for a greater number of websites being restricted.

The impact of censorship on Bing is not limited to users within China. Citizen Lab researchers found that Bing’s censorship of search suggestions, though not search results, was applied to users in the United States and other countries for at least eight months from October 2021 to May 2022.¹⁷ Bing’s censorship of politically sensitive search suggestions in both English and Chinese applied to multiple regions outside China, including the United States and Canada. Suggestions including the names of politically sensitive figures were censored, including those of Xi Jinping, the doctor Li Wenliang who had warned his colleagues about early Covid-19 infections in Wuhan, religious figures including the 11th Panchen Lama Gedhun Choekyi Nyima, and references to the “Tank Man” photographed standing in front of a column of tanks leaving Tiananmen Square on June 5 1989. In response to a May 10 2022 letter addressed Microsoft’s Chief Digital Security Officer, Microsoft communicated to the Citizen Lab that it had discovered and resolved a misconfiguration on Bing which had prevented valid autosuggestions from appearing for users outside China.¹⁸ However, while Microsoft ceased Chinese political censorship of autosuggestions in countries outside of China including the United States, there is no indication that Microsoft has ceased censoring autosuggestions for users of Bing in China.

Microsoft is not the only US company which performs Chinese political censorship. In 2021, Citizen Lab researchers found that Apple applied censorship to product engravings in China.¹⁹ Censored political content included the names of Chinese leaders, Chinese dissidents, and independent news organizations, as well as general terms related to religion, democracy, and human rights. Apple applied these censorship rules not only in China, but in Hong Kong and Taiwan as well. Research findings also indicated that Apple did not fully understand what content they censored. Instead, many censored keywords appeared to have been reappropriated from other sources, including censorship lists compiled by Chinese companies. Since the release of this report, Apple eliminated Chinese political censorship in Taiwan, but has continued keyword-based political censorship in both mainland China and Hong Kong.²⁰

Bing and Apple’s extensive censorship inside China shows that US tech companies cannot introduce services in China without integrating restrictions on expression. Furthermore, our findings show that it is inevitable that such censorship will be applied, either accidentally or otherwise, to users outside of China, including Taiwan and the United States.

Part Three: The Offline Harms of Online Censorship

State-backed restrictions on political and religious expression do not exist in a vacuum. Online censorship is linked to offline harms. Chinese citizens who attempt to access or share sensitive information online do so at risk to their personal freedom. Authorities have jailed Chinese citizens for a range of offenses, including selling software that allows people to circumvent the Great Firewall, making comments in private chat groups, sharing videos of protests, and even posting on social media platforms like X (Twitter) which are blocked in China.²¹ Such cases highlight the severe rights impacts that censorship has on the people of China.

While some Chinese citizens risk detention and even torture for their online activities, state-affiliated actors use these same platforms to launch attacks against opponents of the party-state. As detailed by Citizen Lab researchers, a 2019-2021 harassment campaign nicknamed “HKLEAKS” used websites and social media to distribute personal information about Hong Kong pro-democracy activists.²² Actors involved in the campaign used proprietary websites and social media accounts to publish personal identifiable information about targeted activists. Those connected to the campaign claimed they were members of Hong Kong volunteer committees. However, Citizen Lab researchers uncovered indications that this was a coordinated information operation conducted by professional actors aligned with the Chinese state.

Victims of other online harassment campaigns live outside China. This Commission has previously discussed how the Chinese government silences overseas critics through transnational repression.²³ For years, the Chinese government has used transnational repression to intimidate, threaten, and surveil diaspora members it views as threats.²⁴ Many of these victims are Tibetan.²⁵ Since 2009, Citizen Lab researchers have investigated digital attacks and espionage against Tibetan diaspora communities. These attacks include cyber espionage programs targeting Tibetan institutions, one-click mobile exploits and malware used to install spyware in a target’s phone, and phishing operations conducted against diaspora Tibetan organizations.²⁶

State-backed proxies and online nationalists also harass Chinese, Hong Kong, Tibetan, Uyghur, and other diaspora members on Chinese and US social media platforms. Some of the most vicious instances of digital transnational repression are directed at women.²⁷ As Citizen Lab researchers have documented, Chinese and Hong Kong women activists in Canada have suffered online threats of physical and sexual violence.²⁸ Digital transnational repression has profound consequences for victims and their relatives. Many suffer intense psychological harm, while others self-censor or limit their online activities. Still others have had to contend with state harassment of family members in China, a form of transnational repression known as “coercion-by-proxy.”²⁹

Recommendations

Through state-backed online censorship, the cooperation of Chinese and US technology companies, and domestic and transnational repression, the Chinese state severely restricts the freedom of opinion and expression of people in and outside China. Addressing restrictions on these freedoms requires holding Chinese and US companies responsible for their role in online censorship and supporting victims of digital harassment and intimidation. Therefore, I recommend that the United States government do three things:

One, publicly request that Microsoft, Apple, and other US companies explain how and why they implement political and religious censorship on their platforms in China. Citizen Lab researchers have discovered censorship rules that US companies have implemented on China-accessible platforms and measured the breadth and impact of these rules. However, it is not clear how US companies develop and implement these censorship rules, nor why US companies are willing to censor political and religious content on their China-accessible platforms. Requesting Microsoft, Apple, and other US companies to provide this information would contribute to more informed and effective policies pertaining to addressing the rights and privacy impacts of online platforms and digital technologies.

Two, publicly request that Microsoft explain how political and religious censorship was applied to the search suggestions of users of Bing outside China and what safeguards will ensure this will not reoccur. Citizen Lab researchers discovered that for a period of at least eight months from October 2021 to May 2022 Microsoft’s Bing search engine censored politically sensitive Chinese search suggestions in different world regions, including the United States. It is unclear why Microsoft censored these suggestions and what steps Microsoft has taken to prevent this kind of censorship from reoccurring. Requesting Microsoft answer these questions would deepen understanding of how individuals outside China, including in the United States, are impacted by Chinese state-backed censorship on China-accessible platforms.

And three, train US government officials, including law enforcement and immigration authorities, to recognize digital transnational repression and properly assist victims and their families. Many victims of transnational repression, including digital transnational repression, live in the United States. The United States government has a duty to protect both US and non-US citizens who are victims of transnational repression. Providing protection requires recognizing the severity of the problem. Personnel working in relevant government offices, including US Citizenship and Immigration Services and federal, state, and local law enforcement, should receive training to help them identify both victims and perpetrators of transnational repression. Training should also include learning how to conduct outreach to victims and their families and how to provide appropriate assistance to those at risk of transnational repression. By helping victims of transnational repression in the United States, the US government will demonstrate support for those exercising their freedom of expression and opinion on and offline.

Thank you for the opportunity to testify today. I look forward to your questions and comments.