MENLO PARK, California -- Meta has taken down more than 200 China-based accounts spreading disinformation on social media around the world.
The networks originated in China and targeted social media users in many regions, including in the United States, Europe, India, Tibet, Taiwan, sub-Saharan Africa, Japan, Central Asia and the Uighur community around the world.
They are accused of violating Meta's policy on co-ordinated inauthentic behaviour (CIB) and were taken down before they could build an audience, the Facebook parent company said in its quarterly Adversarial Threat Report published Wednesday (May 3).
"These latest takedowns signal a shift in the nature of the China-based CIB activity we've found with new threat actors, novel geographic targeting and new adversarial tactics," the report said.
"These latest networks experimented with a range of tactics we haven't seen in China-based operations before," it said.
The networks operated on multiple platforms including Facebook, Instagram, YouTube, Twitter, Telegram, PayPal, cryptocurrency, Blogspot, Reddit, Wordpress and Freelancer.com, according to Meta.
Meta said it had removed 107 Facebook profiles and 35 Instagram accounts linked to one network.
"The individuals behind this network posted mainly in English, Russian, Uyghur and Chinese about news and current events in the regions they targeted," Meta said.
Topics included "geopolitics in Central Asia, the impact of the Collective Security Treaty Organisation (CSTO), the Organisation of Turkic States and anti-Russia sanctions on Central Asia", Meta added.
The propaganda also featured warnings against boycotting the 2022 Beijing Olympics, alleged US foreign policy misdeeds in Africa and abuses of migrants in Europe, particularly Muslim refugees.
Other content featured "claims of comfortable living conditions for Uyghurs in China and criticism of politicians in Taiwan".
New behaviour
Meta added that it took down another 50 Facebook profiles and 10 Instagram accounts linked to another China-based network.
"The latest behaviours included creating a front media company in the West, hiring freelance writers around the world, offering to recruit protesters and co-opting an NGO in Africa," it said.
While Meta had taken down some of the accounts, much of the networks' content remains online, AFP reported.
On Twitter, an account named in the Meta report called New Europe Observation shared incendiary content attacking migrants to Europe and lesbian, gay, bisexual, and transgender (LGBT) activists as recently as April 28.
One tweet by the group from last August said it was hiring "part-timers" to attend a protest in Hungary against billionaire philanthropist George Soros, a financier who is frequently a target of far-right conspiracy theories.
Some of its Twitter posts used viral tweets from far-right accounts.
Interspersed with the Europe-focused messaging was content pushing Beijing's line on its policies in China's Xinjiang region.
Beijing has detained more than one million Uighurs and other mostly Muslim, Turkic-speaking inhabitants of the far-western Xinjiang region in a secretive network of detention centres and prisons over the past few years.
Accusations include mass incarceration, forced labour, compulsory sterilisation, systematic rape and the destruction of Uighur cultural and Islamic sites.
Another account that Meta said was run by the network focused more closely on divisive issues in the United States, including police brutality, crime and LGBT rights.
The network was run through a UK-based front company called London New Europe Ltd, Meta said, which company records and Google Maps showed operated out of a nondescript apartment building in northeast London.
The group sought to hire Central Asian freelance writers to produce content, as well as "tried to engage individuals to record English-language videos scripted by the network", Meta said.
Meta said that, while the network took steps to conceal its origins and the identities of those involved, it had found links to a Chinese company called Xi'an Tianwendian Network Technology.
It also operated Monday through Friday from 9am to 5pm China time, with a dip in activity during lunchtime and much less activity on weekends, Meta said.
Chinese company records seen by AFP showed the company's boss had the same name as a Chinese national listed as a director of London New Europe Ltd in the UK.
Xi'an Tianwendian Network Technology did not respond to requests for comment from AFP.
AFP also sought to contact New Europe Observation on Thursday via an email address listed on its Twitter account but received no reply.
Another China-based group primarily focused on India and Tibet named in the report spent $73,000 on Facebook advertising, Meta said.
Twitter and Facebook are officially blocked in mainland China, with users needing banned virtual private networks to use their websites and apps.
Adversarial threats
Meta started its public threat reporting about six years ago when it first shared its findings on a Russian influence operation.
Russia has long been accused of operating "troll farms" in a bid to influence public opinion in the West, with China not having been regarded as advanced in that area.
But Meta said the latest networks it had uncovered signalled China-based operations were becoming more sophisticated.
US authorities also said last month they had charged a group of Chinese Ministry of Public Security officers, saying they had operated a network of social media accounts spreading pro-Beijing messaging.
Around the same time, US law enforcement arrested two men for setting up a Chinese "police station" in New York.
And in February, research firm Graphika revealed it had discovered a China-linked network promoting a fictitious news outlet called Wolf News that used artificial intelligence-generated anchors to spread pro-Beijing talking points.
Meta also removed 40 Facebook accounts run by an Iran-based network targeting primarily Israel but also Bahrain and France.
"This operation ran across multiple internet services -- including Facebook, Twitter, Telegram, YouTube and hacking forums -- where it claimed to have hacked entities in the countries they targeted," Meta said, adding that it could not confirm if the claimed hacks had occurred.
Targets included "news media, logistics and transport companies, educational institutions, an airport, a dating service and a government institution", it said.