Three pitfalls to avoid when analyzing pro-PRC inauthentic messaging

images of robots

Three pitfalls to avoid when analyzing pro-PRC inauthentic messaging

by Hannah Lincoln

Two Six Technology’s Media Manipulation Monitor (M3) tracks thousands of inauthentic Twitter accounts with a history of promoting narratives aligned with the People’s Republic of China (PRC) government’s priorities. After systematically analyzing their activity and reviewing existing scholarship on the topic, we’ve identified three mistakes that are easy to make when analyzing pro-PRC inauthentic activity in the global information environment.

Pitfall 1: Attributing inauthentic activity to the Chinese government without sufficient indicators. 

A good story is not an indicator. On 1 December 2022, a week after anti-lockdown protests broke out in China on 24 November, nearly 200,000 Twitter accounts began to simultaneously promote prostitution services in Chinese cities. 

Some observers attributed the activity to the Chinese government as an attempt to prevent further collective action from spreading by hashtagging city names where protests were forming, thus crowding out conversation about the protests. However, a random sample of 500 participating accounts did not show any indicators of government alignment before 1 December or in the six months after, through May 2023. Furthermore, inauthentic campaigns citing Chinese cities in hashtags to promote prostitution services are actually frequent: M3 captured six instances of such campaigns surpassing 250,000 Twitter posts in one day from January–June 2023, none of which showed pro-PRC messaging or appeared on China’s domestic social media platform, Weibo. 

M3 now assesses that a private sector entity is more likely the sponsor of the 1 December campaign than the PRC government. That campaign—and other similar campaigns on Twitter—likely target PRC-based netizens to sell them services, scam them for money, or steal their data. 

The Chinese language itself is another red herring for PRC-backed inauthentic activity. Chinese-language content from inauthentic accounts is not an automatic indicator of PRC sponsorship, and mis-attribution of this signal can lead to gravely misinformed analysis. M3 assesses that half of Twitter content talking about China in Chinese is anti-PRC.  

  • M3 reviewed a random sample and found that 50% of original Twitter posts from unverified accounts mentioning the phrase “China” in simplified Chinese (“中国”) express anti-PRC views and are likely sponsored by overseas Chinese dissident groups (most frequently showing support for Chinese dissident Guo Wengui). 

Pitfall 2: Assuming inauthentic messaging represents Beijing’s true or only message. 

Imagine you are a Chinese citizen working a low-paid job at your local government office as a social media promoter, managing a slew of fake Twitter accounts that spread pro-China and Beijing-aligned content. As you’re scrolling through Twitter, you see accounts spreading Pfizer vaccine disinformation. Knowing that Pfizer is a US company and that its vaccines are not available in China, you decide to help spread this disinformation. No one told you to spread this message—it just seems like the kind of thing you should be doing to please your bosses.

This hypothetical scenario illustrates how inauthentic accounts can push messages that do not align with Beijing’s official message, in particular if they are run in a decentralized manner. M3 identified at least six instances of the PRC government and inauthentic messaging differing in response to global news events during 2022–2023, one of which included the spread of anti-Pfizer disinformation. Many inauthentic accounts have some freedom to post their own interpretation of what they think they should be saying. Covert messaging should therefore not be equated with official messaging, and analysts should not assume that covert messages represent the central government’s true or only message.

Pitfall 3: Interpreting the Chinese government as a monolith. 

It’s useful to know what department within the government is likely sponsoring inauthentic messaging because that will help identify and rank likely motives. Did inauthentic accounts promote a message to turn US voters against a candidate, to confuse people about human rights abuses in Xinjiang, or to drive up a PRC diplomat’s follower count? Government departments and officials use inauthentic accounts for different reasons. 

The inauthentic accounts that amplify a Chinese diplomat’s posts might be hired by that diplomat’s embassy to increase their reach in the country of residence or to raise that particular diplomat’s profile. Meanwhile, PRC state media and the Central Propaganda Department (CPD) are likely in charge of hiring inauthentic accounts to promote their own respective articles and posts. There is also the People’s Liberation Army (PLA), the United Front Work Department (UFWD), and the Public Security Bureau (PSB), all of whom employ a mix of external and in-house human labor to covertly promote their messages, according to reports. 

Bureaucracies are not known for close, consistent, and rapid coordination across departments, in particular on covert strategies. Analysts should therefore consider what department or, if possible, what government official could be behind an inauthentic message to assess if it represents the PRC’s whole-of-government messaging strategy or one person’s political aims—or (as in most cases) something in between.