Beginning late last week and into Monday, searches in Chinese for major protest hotspots, including Beijing, Shanghai, Nanjing, and Guangzhou, produced a nonstop stream of solicitations, images of scantily clad women in suggestive poses and seemingly random word- and sentence fragments. Many of the tweets reviewed by CNN on Monday came from accounts that had been created months ago, follow virtually no other accounts and have no followers of their own.
The spike in suspected inauthentic behavior followed a deadly fire in China’s Xinjiang province, where at least 10 people were killed amid Covid-19 lockdown restrictions that reportedly hindered first responders from reaching the blaze. The fire, and long simmering frustration over the country’s zero Covid policies, helped spur the rare protests in China.
“It is happening not just around Xinjiang but around any sensitive Chinese issue at the moment,” said Charlie Smith, the pseudonymous co-founder of GreatFire.org, a digital activism group based in China. “Search any city that has seen a rise in Covid cases, or had on-the-street protests on the weekend, and you will see the same thing.”
The apparent suppression campaign by suspected bot accounts represents one of the first major disinformation tests for Twitter since the platform was purchased by Elon Musk. The billionaire has personally vowed to wage war against bots and spammers but has also cut more than half of Twitter’s staff, raising concerns about the company’s ability to combat bad actors in the United States and abroad.
US lawmakers have expressed alarm about Twitter’s alleged vulnerability to foreign exploitation. Moreover, Musk’s ties to China through one of his other companies, electric-vehicle maker Tesla, have raised doubts about his willingness to stand up to the Chinese government.
Twitter, which has cut a substantial amount of its public relations team, didn’t immediately respond to a request for comment.
GreatFire.org, which helps Chinese citizens get around the country’s internet censorship, noted a torrent of “dating” spam tweets appearing on Friday tagged with “Urumqi,” the capital of Xinjiang. The flood of spam tweets is still ongoing, Smith told CNN on Monday.
Pornography and sex-related sites were among the first to be censored by China when it began its internet crackdown years ago, Smith added, making it less likely that the spam tweets advertising sex services are the work of random, private individuals.
Twitter is officially blocked in China, but estimates of the number of Twitter users in China have ranged between 3 million and 10 million.
On Sunday evening, Alex Stamos, director of the Stanford Internet Observatory and a disinformation researcher, elevated an independent researcher’s findings that Stamos said “points to this being an intentional attack to throw up informational chaff and reduce external visibility into protests in China.”
The other researcher’s self-described “quick and dirty analysis” of the location-focused searches suggested a “significant uptick” in recent tweets containing ads for escorts, pornography and gambling.
Stamos, who previously worked as the chief security officer at Facebook, later tweeted that the apparent disinformation campaign has convinced him to seriously consider leaving Twitter. “We are rapidly approaching the point where any political discussion will be dominated by organized influence teams and more lighthearted topics by spam,” he said.
Musk has pushed back on suggestions that his ownership of Tesla, which is heavily invested in China, may give the Chinese government “leverage” over Twitter. In June, prior to completing his purchase of the social media company, Musk told Bloomberg News that “as far as I’m aware,” China does not attempt to interfere with the free speech of the US press.
But for years, social media companies including Twitter have highlighted actual and multiple examples of foreign influence operations on social media. The recent layoffs and resignations at Twitter — which have directly affected the teams responding to Chinese influence campaigns, a former employee told The Washington Post — have further reduced the company’s ability to meet those challenges.
It’s also unclear to what extent China may have visibility into Twitter’s service and internal systems. Earlier this year, Twitter’s former head of security told the US government in a whistleblower disclosure that the company is extraordinarily vulnerable to foreign exploitation. The whistleblower’s testimony claimed the FBI had warned the company this year that at least one agent working for the Chinese government was on Twitter’s payroll.
The claim has alarmed US policymakers. Last week, Sen. Chuck Grassley, the top Republican on the Senate Judiciary Committee, wrote to Musk asking him to review Twitter’s security for insider threats and to brief congressional staff on the matter.
