Flood global social media with fake accounts used to promote an authoritarian agenda. Make them look real and increase the number of your followers. Search online critics of the state and find out who they are and where they live.
China’s government has launched a global online campaign to polish its image and lower allegations of human rights abuses. Much of the initiative takes place in the shadows, hidden behind networks of robots that generate automated posts and hard-to-find online profiles.
Now, a new set of documents seen by The New York Times reveal in precise detail how Chinese authorities use private companies to generate on-demand content, attract followers, identify critics and offer other services for information campaigns. This operation is increasingly taking place on international platforms such as Facebook and Twitter, which in China are blocked by the government.
The documents, which were part of a tender for contractors, offer a rare glimpse of how China’s extensive bureaucracy works to spread propaganda and shape public opinion on global social media. They were taken down after the Times contacted the Chinese government to talk about them.
On May 21, a branch of the Shanghai police posted an online advertisement asking for proposals from private contractors for what Chinese authorities call “public opinion management”. Authorities have relied on tech companies to help them monitor domestic social media and actively shape public opinion through censorship and the dissemination of fake posts across the country. Only recently have policymakers and the opinion management industry turned their attention beyond China.
Shanghai police wanted to create hundreds of fake accounts on Twitter, Facebook and other major social networking platforms. The police department emphasized that the task was urgent, suggesting it wanted to release the bills quickly to influence discussions.
Robotic account networks like the ones the Shanghai police want to buy have fueled a surge in pro-China online traffic over the past two years. Sometimes posts from these networks on social platforms reinforce official government reports with “likes” or reposts. Others attack network users who criticize government policies.
The police department wanted an upgrade in sophistication and potency: a series of accounts with real followers that can be turned to government purposes when necessary.
The request suggested that law enforcement authorities understand the need for strong engagement with the public through these rental profiles. Deeper engagement lends credibility to fake profiles at a time when social networking companies increasingly take down accounts that look fake or coordinated.
The two robotic networks that have been linked to the Chinese government stand out for their lack of engagement with other accounts, according to misinformation experts. While they can be used to deceive people and reinforce the number of likes on government posts, most of these automated accounts have little individual influence as they have few followers.
In their post, authorities used a common phrase among Chinese internet police, which means tracking the real person behind a social network account: “touching the ground”.
With increasing frequency, the country’s internet police have hunted down and threatened world wide web users who voice their opinions. At first, its agents focused on social media platforms. In 2018, they began a new campaign to stop Twitter users in China — account holders who had found ways around government blockades — and force them to delete their accounts.
The campaign now extends to Chinese citizens living outside China. The document explains how the Shanghai police intend to discover the identity of people behind certain accounts and trace their users’ connections to the mainland. Police officers can then threaten their families in China or detain account owners when they return to the country, to force critics to delete posts or even entire accounts.
In previous information campaigns in China, robotic accounts were used to add an unrealistic number of likes and retweets from government and state media posts. The traffic flow created can increase the likelihood that posts will be shown by recommendation algorithms on many social networking sites and search engines.
As China’s advertising campaigns developed abroad, they relied more on visual media. Authorities were looking for a company to maintain and mobilize fake accounts and also to generate original content. Demand for videos is high.
Another document seen by the Times shows that the same local branch of Shanghai police purchased video services from another company in November. Police asked the provider for at least 20 videos a month, as well as their distribution on domestic and international social media. The document referred to the task as producing an original video that would be used to fight the “battle of public opinion”.
Earlier this year, an analysis by the Times and ProPublica showed that thousands of videos portraying Uighur ethnic minority members living happily and free were a key part of the disinformation campaign that Twitter ended up attributing to the Chinese Communist Party. When Twitter took down the network behind these posts, it canceled accounts linked to a company it said had helped make advertising videos. A Twitter spokesman declined to comment.
Three weeks after the Shanghai police department’s request was released, a company called Shanghai Cloud Link won the bid, according to the documents. In its proposal, the company said it had only 20 employees. According to its founder Wei Guolin’s Linkedin page, the company works with multinational firms and offers “digital governance” and “smart cities” services.
Wei did not respond to a request for comment. The Shanghai Pudong Public Security Department did not respond to a faxed request for comments.
Jobs like the one announced by Shanghai Cloud Link are probably just the tip of the iceberg. Local governments and police across China have posted requests for similar services to influence social networks abroad, but generally in vague terms. Sometimes details are revealed. In 2017, for example, Inner Mongolia police purchased software that allowed government trolls to post directly to various social networking sites, inside and outside China, according to documents seen by the Times.
In another case, a contractor had downloaded hundreds of credentials to access Facebook’s public feed, allowing it to collect data on who commented on the posts and when. Facebook did not make immediate comments.
Shanghai Cloud Link’s winning bid offers a window into how much these disinformation services can cost.
In many cases, technology companies try to sell hardware and software directly to Chinese authorities. In this case, Shanghai Cloud Link’s proposal suggested a new service-based model, in which authorities pay a monthly fee, a kind of subscription for manipulating social networks.
.
