Beijing’s Computational Propaganda Goes Global: The Significance of China’s Debut as a Disinformation Actor

Only three weeks ago, Twitter made history with its announcement that the government of the People’s Republic of China (PRC) was spreading disinformation to discredit the ongoing protests in Hong Kong opposing a new proposed law to extradite criminals in Hong Kong to mainland China. Acting on leads passed on by Twitter, Facebook subsequently followed suit with a more circumspectly worded announcement of a takedown of a smaller, related China-backed disinformation campaign. For disinformation experts and researchers, the uniqueness of the occasion was not lost — this is the first time social media platforms have formally named China as an information operations (IO) actor. For the many who don’t spend their days analyzing social media posts or tracking disinformation, however, the announcement may not have stood out. Many of you may be wondering — what’s the big deal? Didn’t we know China was doing this kind of stuff anyway?

The answer is a somewhat complicated no. While there has also been ample evidence of pro-China activity on social media in years past, much of it seemingly coordinated, there has not been attribution of Chinese state-sponsored disinformation on international platforms such as Twitter or Facebook. There have however been five peripheral, closely related activities that are worth noting.

• Coordinated pro-China activity: evidence of coordinated support for China and state narratives on social media has grown in recent years. In 2019 alone, we have seen mass harassment of a Tibetan student in Canada on Instagram, mass downvoting of content critical of the PRC on Reddit, and inauthentic support for the pro-China mayor Han Kuo-yu, who recently clenched the presidential nomination for Taiwan’s right-wing party, the KMT. In late 2015 and early 2016, French journalist Ursula Gauthier received vitriolic harassment for her reporting on the Xinjiang region, which was ultimately a prelude to her expulsion from the country. Even farther back, in 2012, cybersecurity expert Brian Krebs noted the presence of bots spamming the hashtag #Tibet with pro-China messages. While much of this activity has hinted at coordinated or inauthentic behavior, it has been far from revealing any state involvement.
• Domestic state-sponsored information operations: China is known to target Chinese citizens on domestic platforms with online propaganda operations through the 50 cent army (五毛党) — users the government pays to post pro-PRC comments online.
• International, state-sponsored hacking operations: China has shown itself willing to engage in traditional cyber warfare to gain intelligence or help allies in elections, most notably in the Chinese hacking of Cambodian electoral infrastructure, as well as the GhostNet spyware operation targeting Tibetan institutions in 103 countries, which is suspected to have been undertaken by the Chinese government.
• Offline influence operations: the use of sharp power tactics, purchase of Taiwanese media companies by pro-China companies and buying positive coverage of China in local Taiwanese newspapers have all been vectors of offline influence in recent years.
• Growing influence of Chinese state-sponsored outlets on Facebook: Out of the six media outlets with the biggest followings on Facebook, five are Chinese state-sponsored outlets. The PRC has also substantially invested in state media messaging on Facebook in recent years.

Despite these closely related operations, until weeks ago, there hadn’t been attributed evidence that China spread disinformation on Western platforms through fake accounts and astorturfing efforts similar to those undertaken by Russia, Iran, Venezuela and Bangladesh.

The announcement indicates China’s willingness to use these same underhanded tactics on Western platforms when matters of domestic or foreign policy reach a breaking point. China’s foray into online information operations also further validates a hypothesis advanced by US intelligence chiefs and former US Ambassador to NATO Victoria Nuland: that Russia’s disinformation campaign targeting the US in 2016 and before has inspired other countries to undertake similar efforts. As other experts have noted, this also raises the probability of China engaging in similar information operations targeting the US presidential race in 2020.

The Dire Strait: Looking Ahead to Taiwan

One of the most immediate questions raised by Twitter’s announcement centers on Taiwan. Taiwan will be electing a new president in national elections in January 2020. The island’s politics and relationship with China are one of the most important foreign policy issues to mainland China. Taiwan has an awkward diplomatic status — while it functions as an independent country with one of Asia’s most free press spheres, China claims dominion over the island due to its never having formally declared independence after the end of the Chinese civil war in 1949. With tensions high in Hong Kong, and coverage and criticisms of China’s surveillance state and persecution of Muslims in its Northwest region of Xinjiang, China may feel it cannot afford to let Taiwan be a runaway train. Taiwan has one of the highest rates of Facebook penetration on Earth: 97% of internet users in the country use the platform, making Facebook a uniquely effective attack surface in Taiwan for 2020.

Successful Chinese disinformation on Western platforms would be particularly valuable given the appearance of a pro-Chinese right-wing politician, Han Kuo-yu (韓國瑜), who recently won an upset victory in the mayoral race in Kaohsiung, Taiwan’s second largest city, in November 2018. Seemingly coordinated and inauthentic activity promoting Han occurred that year, much of it emanating from profiles using a written form of Chinese only used in mainland China.

Upcoming Analyses

Twitter has laudably released attributed data from China’s IO campaign targeting Hong Kong, allowing external researchers and experts to analyze and investigate independently. What interesting patterns show up in this data? What are the main narratives pushed in the data? What operational similarities does China share with other actors like Russia and Iran? What role did bots play in this campaign?

The Digital Intelligence Lab will be producing a public analysis of the Chinese disinformation dataset released by Twitter to help shed light on these questions. It is our hope that through collective analysis of this public data, we can begin to understand China’s modus operandi when conducting influence operations online, and help inform technical and social solutions to the ever-evolving problem of online disinformation.