Facebook today shared the details of an attack campaign that used its platform as part of a broader operation to spy on Uyghur Muslim journalists, activists, and dissidents around the world. Officials say a Chinese group is responsible for the advanced attack.
This group used Facebook to create fake accounts, which have now been removed, and distribute links to malicious websites and iOS and Android malware. Attackers used the social platform to target Uyghurs from Xinjiang, China, who now live in the United States, Turkey, Kazakhstan, Syria, Australia, Canada, and other countries, the company reports.
News of the attack arrives the same week that the US, Canada, European Union, and United Kingdom imposed sanctions against Chinese officials for "serious human rights abuses" against Uyghur Muslims, who have been the targets of mass detention in China.
This campaign started in 2019 and affected at least 500 targets; however, Facebook says this only accounts for parts of the attack that somehow touched the platform. Most of the attack activity did not, says Nathaniel Gleicher, head of security policy for Facebook.
Attackers built malicious third-party websites that used lookalike domains for popular Uyghur and Turkish news websites; they also seem to have compromised legitimate sites that Uyghurs visit as part of watering-hole attacks. Some sites held malicious code similar to previously reported exploits that installed Insomnia iOS malware on devices.
To distribute these malicious links, the attackers used fake Facebook accounts to pose as reporters, students, human rights advocates, and other Uyghur community members to establish trust with their victims and trick them into clicking on the malicious links.
The group was careful to hide their activity by only deploying the iOS malware when a target met specific technical criteria, such as IP address, operating system, browser, and country and language settings, says Mike Dvilyanski, Facebook's head of cyber-espionage investigations. This activity was highly targeted and designed to collect people's data.
Facebook also found websites designed to resemble third-party Android app stores, where attackers put fake apps that might appeal to Uyghur targets. These included a keyboard app, prayer app, and dictionary app, all of which contained the ActionSpy or PluginPhantom Android malware strains.
Analysis revealed two Chinese companies, Beijing Best United Technology and Dalian 9Rush Technology, are behind some of the Android tools. Facebook notes FireEye research contributed to their assessment.
"FireEye uncovered an operation targeting the Uyghur community and other Chinese speakers through malicious mobile applications that were designed to collect extensive personal information from victims, including GPS location, SMS, contacts lists, screenshots, audio, and keystrokes," says Ben Read, director of analysis for Mandiant Threat Intelligence, in a statement, noting the operation FireEye has been following has been active since 2019.
Facebook did not directly attribute this attack to the Chinese government. While it can see the geographic attribution, officials say, it can't prove who is behind the operation.
"Our industry peers have been tracking parts of this activity as being driven by a single threat actor broadly known as Earth Empusa, or Evil Eye, or PoisonCarp," Gleicher and Dvilyanski write in a blog post on the attack. Facebook's investigation has confirmed the activity it has disrupted so far closely aligns with the first two. While PoisonCarp shares some of the techniques, its analysis shows this is a separate cluster of activity.
Facebook has blocked the sharing of these malicious domains on its platform, removed the attack group's fake accounts, and notified people believed to be targeted. It's sharing its findings today to expand disruption efforts, as it expects attacks to continue.
"We saw this activity slow down at various times, likely in response to our and other companies' actions to disrupt their activity," the post states.