A pro-Chinese digital propaganda network is attempting to discredit COVID-19 vaccines distributed in the United States, according to a new, independent report by the social media research firm Graphika.

The influence campaign is part of a propaganda network that has bought and repurposed thousands of fake YouTube, Twitter and Facebook accounts to defend China’s response to the pandemic — and to criticize the U.S. While the network’s earlier efforts failed to attract an organic audience, its recent maneuvers are more sophisticated and are gaining momentum, having been amplified by Chinese diplomats and influencers in Latin America, Pakistan and Hong Kong, Graphika found in a report released to the public Thursday morning.

“It’s trying to adapt to survive,” said Ben Nimmo, one of the report’s authors and the head of investigations at the New York City-based Graphika. “There have been this handful of occasions where it has gotten some traction from amplifiers.”

While much of China’s COVID-19 messaging is in Mandarin and is targeted at bolstering domestic support, its English-language propaganda also serves a function, said Nick Monaco, a senior China analyst for the digital consulting firm Miburo Solutions. Highlighting America’s problem-plagued response to the pandemic can be an effective diplomatic tool for the Chinese Communist Party, said Monaco, who was not involved in Graphika’s new report.

“It’s an opportunity to promote the Chinese authoritarian model as more efficient and more effective,” Monaco told FRONTLINE.

Graphika’s investigation, which spanned the last six months on YouTube, Twitter and Facebook, builds on the firm’s three previous reports on the network, which, in addition to COVID-19 messaging, has promoted content related to arms control, economic development in China and political conflict in the U.S.

The network first emerged in 2018, using fake Twitter and YouTube accounts to spread criticism of exiled billionaire and Steve Bannon associate Guo Wengui, as well as of the Hong Kong pro-democracy movement. But its activity has increasingly focused on the U.S. over the last year, as the Trump administration ramped up its rhetoric against China and the country’s Sinovac vaccine suffered disappointing clinical trial results in Brazil.

In February 2020, as the Chinese Communist Party faced growing criticism for its lack of transparency regarding COVID-19 — a period explored in the new FRONTLINE documentary China’s COVID Secrets — the network shifted its focus to defending China’s response to the pandemic. In June, the accounts began posting crudely edited videos attacking America’s failing efforts to contain the pandemic.

While many videos contained slanted takes on legitimate news stories, others promoted false claims. One video posted by network accounts in November, titled “Comparison of China and the United States in the fight against COVID-19,” said that the U.S. pandemic strategy included drinking hand sanitizer. And two videos in January 2021 suggested without evidence that the Pfizer-BioNTech vaccine currently being distributed in the U.S. is unsafe. One video, titled “Vaccines Will Not Get America Out of This Mess,” said: “The safety of the [U.S.-developed Pfizer-BioNTech] vaccine was in doubt, but it was quickly approved.” America’s second COVID vaccine, developed by the Massachusetts biotech company Moderna, was not mentioned.

The ramping up in messaging over the last year has corresponded with a move toward more sophisticated tactics. While the network seems to have had tens of thousands of bot accounts at its disposal since 2018, each had relatively few followers and were frequently taken down by platforms. In 2020 alone, YouTube deleted more than 20,000 of the network’s accounts, according to Google’s Threat Analysis group. A Twitter representative told FRONTLINE the platform has taken large-scale action against network accounts, including mass removals in August 2019 and June 2020. Facebook took down a wave of accounts in September 2019 and has since worked with Graphika to detect and delete new ones.

Since mid-2020, however, the network has improved the fake accounts’ ability to imitate humans.

In one case, a Twitter account that had been dormant since 2009 added a profile photo and began tweeting the network’s videos in January 2020, when it also began replying to political figures and celebrities. Foreign Ministry Deputy Spokesman Lijian Zhao retweeted the account, as did China’s ambassadors to Iran, the Dominican Republic, France and Panama. 

The network’s traction with real humans represents a shift, Nimmo said. But even with recent upgrades, Nimmo said the operation is still relatively unconvincing, with the video voiceovers often including grammatical errors — pronouncing “U.S.” as “us,” for example — and idioms that do not translate into coherent English.

“Yes, it’s a very large-scale network; yes, it’s spammy; yes, it’s starting to get signs of breakout in some areas, in individual cases,” Nimmo said. “But fundamentally they’re still not very good at what they do.”

Many of the spam accounts appear to be purchased from “like farms” that sell fake engagement to influencers, the report said — for example, Facebook pages created in Bengali that later began posting in Chinese and amplifying the network’s videos. Graphika used multiple behavioral cues to determine whether accounts were likely part of the network, including posting the same pro-regime videos, interacting with other demonstrably fake accounts created around the same time, and beginning to post network content after having been dormant for years.

Graphika can’t determine who is behind the accounts, Nimmo told FRONTLINE. Disinformation researchers from the Stanford Internet Observatory and the digital consulting firm Miburo Solutions also told FRONTLINE it can be difficult to determine whether Chinese influence campaigns are organized by the ruling Communist Party or by nationalist citizens.

The Chinese Embassy in Washington, D.C. did not return a request for comment before publication.

Even with limited results so far, the network’s scope — including the bulk purchase of fake accounts and the posting of multiple unique videos per day — suggests an organized and funded effort, Nimmo said. “It’s hard to imagine something of that scale would be a bunch of amateurs.”

The Graphika report comes at a sensitive moment for China’s government, as a World Health Organization team on the ground in Wuhan is conducting its long-awaited investigation into the origins of the pandemic. 

The U.S. State Department did not comment on the specific content of Chinese propaganda covered in Graphika’s report but told FRONTLINE it supports a “robust and clear” investigation of how the pandemic began.

“We do not know, and cannot know, the full origins of the virus and its spread until we have full transparency from the [People’s Republic of China]. It is imperative to have transparency on the early days of the pandemic, including to identify important lessons about how future pandemic emergence events can be averted,” a State Department spokesperson wrote in a statement to FRONTLINE.

In addition to Graphika, others have found evidence of widespread disinformation seeded by Chinese sources. Reports from the European Union and the Stanford Internet Observatory have also documented digital messaging designed either to praise the Chinese government’s pandemic response or to discredit COVID-19 response efforts in the United States dating back to February 2020.

FRONTLINE found that the Global Times, a newspaper based in Beijing and operated by the Chinese Communist Party, has posted 24 digital stories mentioning the unfounded conspiracy theory that COVID-19 actually began in Fort Detrick, a U.S. military base in Maryland that studies infectious disease. The earliest was published in March 2020; the most recent was published this week. The claim has also been amplified by Chinese government officials, including a foreign ministry spokesperson who in March tweeted a link to a conspiracy website alleging that the virus originated in the U.S.

In March 2020, ProPublica found that a network of thousands of fake and hacked Twitter accounts had begun posting China-friendly messages about the burgeoning pandemic. One account impersonating the U.S.-government-funded Radio Free Asia touted China’s donation of protective equipment to Italy. The accounts were linked to a company on the payroll of Chinese state media, ProPublica reported.

China has also pointed its digital propaganda apparatus at Taiwan, whose independence is not recognized by the Chinese government. An August 2020 report from Graphika and the think tanks Institute for the Future and the International Republican Institute found that social media accounts based in China spread misinformation discrediting Taiwan’s response to the pandemic.

In one Twitter post, an account imitating the Associated Press used a photograph of victims in body bags from the 2006 Indonesia earthquake to claim that Taiwan was concealing COVID deaths. Other accounts, pretending to be from Taiwan rather than the mainland, used the hashtag #TaiwanPneumonia to suggest the virus originated there, the report found.

“A lot of this stuff has been among the more malicious disinformation I’ve seen,” said Monaco of Miburo Solutions.

---