HONG KONG — A coordinated campaign to suppress online posts about Chinese protests against “zero-Covid” controls has flooded Twitter with sexually explicit images, internet researchers say, echoing previous efforts to prevent information unflattering to the ruling Chinese Communist Party from spilling onto the global internet.
Though the spam campaign has been underway for several weeks, it ramped up significantly late last month after protests broke out in major Chinese cities like Beijing, Shanghai and Guangzhou, said Charity Wright, an intelligence analyst at the cybersecurity company Recorded Future who specializes in China.
Twitter searches for the names of those cities in Chinese characters returned a wave of spam posts including sexually explicit images of women instead of protest-related information, according to Wright and other observers. The account names have a pattern of a Western female name followed by a string of numbers, which “often indicates they are created in an automated fashion,” Wright said.
Nisos, a digital security company based outside Washington, said in a report for NBC News that it was able to track a “dramatic increase in inauthentic Twitter bot account activity” starting around Nov. 24, when the protests began to spread across China.
“The use of Simplified Chinese city names with unrelated posts to overwhelm search results aids in disrupting communication and coordination between protest networks as well as preventing information regarding protests from reaching western audiences,” Nisos wrote in its report.
Though it is uncertain whether the campaign was directly conducted or sponsored by the Chinese government, Wright said it was hard to imagine who else would be able to conduct such a large-scale campaign.
Liu Pengyu, spokesperson for the Chinese Embassy in Washington, denied the allegation, saying it “has no factual basis.”
“Some U.S. officials, members of Congress, media and institutions have been seeding and spreading a large amount of disinformation against China without producing any evidence,” he said in an email on Monday.
Other researchers have also tracked the spam campaign. Kenton Thibaut, a resident fellow at the Atlantic Council’s Digital Forensic Lab who focuses on China, said the campaign was particularly noticeable as it seemed Twitter’s ability to counter it had been hampered by recent layoffs under new owner Elon Musk. The layoffs were especially severe in teams responsible for content moderation and fighting misinformation.
Such spam campaigns, which have also been observed in relation to political events in countries like Russia and Syria and were once viewed as an advanced form of digital propaganda, are now considered relatively unsophisticated.
Before Musk’s layoffs at Twitter, “this kind of stuff could be handled pretty quickly and the information environment could be a little bit clearer for true information,” Thibaut said. “But because of that it’s really allowed this kind of stuff to come in and muddy the waters.”
Chinese authorities say their “zero-Covid” controls are necessary to save lives in an undervaccinated population that has barely been exposed to the virus three years into the pandemic. But they have moved to ease some restrictions after tamping down the protests, which were the biggest show of public unrest the country has seen in decades.
Videos of the protests quickly spread on Chinese social media, with online censors struggling to keep up. Many of them also made their way onto Twitter and other social media platforms like Facebook and Instagram, all of which are blocked in China but some Chinese users access via virtual private networks.
There are no firm numbers on the scope of the spam campaign, but Wright said she has observed thousands of Twitter posts an hour in some cases. This type of flooding under popular hashtags makes it more difficult for journalists and Chinese living abroad to learn about the public dissent, she added.
“If an audience is going onto a social media platform to search for what’s happening in Beijing, or what’s happening in China, they’re not looking for this kind of content, so they’re more likely to scroll through and avoid that content,” Wright said. “So essentially, it can be very effective as a method for driving out the truth, the real videos, and the real content about the protests.”
Spam campaigns have accompanied online discussions of sensitive topics related to China for years. After the 2016 presidential election in Taiwan, a self-ruling island Beijing claims as its territory, winning candidate Tsai Ing-wen’s Facebook page was flooded with tens of thousands of pro-China comments. In 2019, Twitter said it had suspended 936 active accounts originating in China that appeared to be part of a “coordinated state-backed operation” aimed at undermining anti-government protests in the Chinese territory of Hong Kong.
Wright noted that the current spam campaign has been targeted primarily at Chinese hashtags, with English hashtags left relatively untouched.
“They’re mostly focused on the Chinese audience in these social media platforms,” she said.
U.S. security researchers have, however, taken note. Alex Stamos, director of the Stanford Internet Observatory and the former chief security officer at Facebook, said on Twitter in late November that the spam campaign might be what finally pushes him to leave the platform.
“We are rapidly approaching the point where any political discussion will be dominated by organized influence teams and more lighthearted topics by spam,” he wrote.
Marie Brockling reported from Hong Kong. Jason Abbruzzese and Kevin Collier reported from New York.