Sharing screenshots of sensitive content in China has just become more risky.
More Chinese social media services are putting hidden watermarks on screenshots that make the images traceable no matter where they are shared, a feature that an analyst said could prevent the spread of censored content.
Some Chinese internet users recently discovered the covert watermarks on screenshots taken on Zhihu, a question-and-answer site similar to Quora. By tweaking the colors of the screenshots, the users found strings of numbers plastered across the page. Some suspect it is information that could be used to identify who took the screenshots.
It follows the example of Douban, an online forum that introduced a similar feature in February. The platform allows administrators of message groups to opt in to the feature, which embeds subtle watermarks in screenshots that contain information of who took the screenshot and its origin, ostensibly to prevent users from lifting others’ content.
But Eric Liu, an analyst tracking Chinese censorship for the U.S.-based China Digital Times, said Chinese social media companies are using such techniques, called steganography, to trace sensitive content to their source and block accounts that share the information on banned foreign platforms.
“It’s a method to stop censored content from spreading,” Liu told VICE World News. “If censored content goes viral on social media beyond the firewall, it would put immense pressure on domestic platforms.”
Zhihu did not immediately respond to a request for comment.
Digital watermarking has for years been used for purposes such as copyright protection and video authentication, and could come in the form of hidden messages or imperceptibly altered pixels. As early as 1997, Playboy considered putting digital marks on its photos to track pirated copies on the internet.
Many office printers also put tracking codes, invisible to the naked eye, on printed documents that associate them with the printer’s serial number. This is how the U.S. National Security Agency in 2017 identified a contractor who leaked classified documents to The Intercept, which posted photos of the documents online.
The introduction of hidden watermarks comes as China further tightens its control over the internet and squelches online dissent. China has strictly required all internet users to register with their real names since 2017. Many social media platforms, which are tasked with censorship, require users to verify their identities with biometric features such as facial recognition.
While most foreign social media sites, including Twitter and Facebook, are blocked by China’s firewall, authorities have sought to track down residents that criticize the Chinese government on these platforms. At least 2,300 residents in China have been punished and even sentenced to jail for their online remarks, including tweets, according to a crowd-sourced database.
More recently, Beijing was outraged by an emerging online campaign to take screenshots on Chinese social media and translate them that highlights rising nationalism in the country on Twitter and Reddit.
State outlets have accused the campaign, dubbed the Great Translation Movement, of smearing the country by cherry picking the most extreme sentiments. Some users suggested the hidden watermarks are introduced to catch the volunteer translators.
Zhihu’s feature, which was first noted by users on Sunday, appears to have been rolled back by Tuesday.
But many Zhihu users were not relieved. “The space for discussion within the domestic internet has never been this dire, a slight offense could get your post deleted, your account suspended and result in a trip to the police station,” one user wrote in a thread discussing the new feature. “It is why more refugees are gathering on foreign platforms, where they can speak relatively freely. But this makes those who want to impose control uncomfortable.”
“I guess other domestic platforms would soon follow suit,” they added.