On Saturday night, WeChat began to show videos of protestors in Shanghai. They would only stay up for a few minutes, chanting about freedom and removing COVID-19 limitations, before they were censored.

Elliot Wang (26-year-old Beijinger) was stunned.

Wang said that he started refreshing frequently, saving videos and taking screenshots of everything I could before it was censored.” Wang only consented to have his English name used in the quote, fearing government retaliation. “A lot of my friends were sharing videos of the protests at Shanghai. They would be taken down quickly, so I also shared them.

Wang was able see the outpouring of grievances and how it highlights the cat-and mouse game between the millions of Chinese internet users, and the country’s massive censorship machine.

The Chinese authorities have tightened their grip on the internet by using a multi-layered, complex censorship operation. This blocks access to nearly all foreign news sources and blocks keywords that could be politically sensitive or harmful to the rule of the Chinese Communist Party. Usually, videos of protests or calls to action are deleted right away.

Images of protests spread quickly on WeChat (a popular Chinese social networking site used by more than 1 billion people) in the wake a fatal fire in Urumqi, northwestern China, Nov. 24, 2015. Many believed that residents were prevented from fleeing the flames by lockdown measures, which the government denies.

Han Rongbin, associate professor at the University of Georgia’s international affairs department, said that the sheer number of Chinese users posting their frustrations on the Chinese internet, along with the ways they tried to evade censors caused a short period of government censorship.

Han stated that censors take time to analyze what is going on and add it to their portfolio of censorship. Therefore, the government must learn how to effectively conduct censorship.

2020 saw the death of Li Wenliang (COVID-19), a doctor who had attempted to warn others about a virus similar to SARS. This sparked outrage and anger at the Chinese censorship system. Before censors deleted posts, users left criticisms for hours.

Chinese internet users used humor and metaphors to share critical messages, even as censors removed posts about the fire.

“Chinese netizens are always very creative because every idea that is successful once will be discovered again by censors,” Liu Lipeng, a critic of China’s censorship practices, said.

Chinese users began posting images of white paper sheets on the internet, Liu said, as a silent reminder that they were not allowed to post words.

Others wrote sarcastic messages such as “Good, good, sure sure right all right yes yes,” or used Chinese homonyms in order to call for President Xi Jinping’s resignation. These included “shrimp moss,” which sounds similar to the words “step down” and “banana pear,” which have the same initials of Xi.

Within days, however, censors began to allow images of white papers. Chauncey Jung, a policy analyst, stated that they would have used a variety of tools. He previously worked as a policy analyst for many Chinese internet companies.

Jung explained that the majority of content censorship is not performed by the state. Instead, it is outsourced to content moderation operations on private social media platforms. These operations use both humans and AI. While some censored posts may not be deleted, they can be made visible to the author or removed from search results. Posts containing sensitive key phrases might be published after being reviewed in some cases.

On Thursday, a Weibo search for “white paper” turned up mostly posts critical of protests. There were no images of white paper sheets or people holding white paper at protests.

Although it is possible to connect to the global internet via China using virtual private networks, these systems are illegal. Many Chinese internet users only have access to the domestic internet. Wang does not use a VPN.

Wang said, “I think I can tell all the mainlanders of my generation that we’re really excited.” “But we are also disappointed that we can’t accomplish anything. They keep censoring, deleting, and even publishing fake accounts to praise cops.

However, the system is strong enough to prevent many users from seeing them. Carmen Ou, a Beijing resident, didn’t initially notice the protests that broke out in China this weekend.

After using a VPN to access Instagram, Ou was only able to learn about the protests.

She said that she tried to look at her WeChat feed, but found no mention of protests. “If it weren’t for a VPN, and access to Instagram. I might not have discovered that such a massive event had occurred.”

Han, an international affairs professor, stated that censorship doesn’t need to be perfect in order to be effective.

He said that “Censorship could be used to stop a large enough population from accessing critical information to mobilize it.”

It is difficult to tell government spam from regular spam because of China’s opaque approach in reducing online dissent.

If you search Twitter for Shanghai or other Chinese cities using Chinese words, you will find protest videos and a flood of new posts featuring racy images of young women. Researchers suggested that the state could be trying to drown out news about protests by creating “not safe for Work” content.

According to Stanford data architect David Thiel, while there was a preliminary analysis by Stanford Internet Observatory that found lots of spam, it did not provide any “compelling evidence” that the spam was intended to suppress information and dissent.

Thiel stated in an email that “I’d be skeptical of anyone who claims clear evidence of governmentattribution.”

Twitter searches for protest-related terms such as “Urumqi Middle Road Shanghai” produced mostly posts about the protests.

Cyabra, an Israeli data analysis company, and another research group who shared analysis with AP stated that it was difficult to discern between a deliberate attempt by Chinese diaspora protesters to drown out information and a regular commercial spam campaign.

Twitter did not respond to a request for comments. Since Elon Musk, a billionaire, took control of Twitter in October, it has not responded to media inquiries. It also cut back a lot of its workforce, including those responsible for moderating spam and other content. Musk frequently tweets about how he is enforcing and enforcing new Twitter Content Rules, but has not commented on recent protests in China.