A Legal Chinese VPN? The Next Phase in Internet Censorship
I’m Jordan Schneider, Beijing-based host of the ChinaEconTalk Podcast. In this newsletter, I comment on and translate articles from Chinese media about tech, business, and political economy. If you’ve been forwarded this email,
Happy new year everyone!
Two quick updates on stories we’ve been following. First, one farm outside Harbin in northeast China fought back against gangs using drones to spread African swine flu by using jamming devices so strong they interfered with flight navigation.
Also, UCommune’s IPO prospectus dropped a few weeks back. The company is gross margin negative and without new funding will likely run out of cash in Q2 2020. In its prospectus, under risk factors, the firm writes that “failure to generate positive cash flow from operations may adversely affect our ability to raise capital for our business on reasonable terms, if at all.” What’s more, “Our business will require [a] significant amount of working capital to support our growth.” I wish them all the best squaring that circle.
This week’s piece comes from Peter Hansen, a precocious recent grad who studied Chinese and computer science at UMD.
Chinese internet users may soon be able to access Twitter, Facebook, Youtube, and Google so long as they use the new Kuniao (酷鸟) browser which advertises “free, convenient, fast, and safe” connections beyond China’s Great Firewall. The software was released in an invite-only beta in November and was quickly swamped with eager users. Due to bureaucratic infighting, it’s unsurprising that such a tool was taken down after a few days. But in the coming years, such software could become widespread. It represents the culmination of trends in Chinese internet censorship stretching back almost a decade.
Did this browser’s entrance herald the end of censorship? Of course not. In reality, this browser’s appearance did not represent the removal of barriers but rather their evolution. In a tradeoff familiar with those outside the Great Firewall, it offered greater services in exchange for more personal information. Users would be required to register their phone numbers in order to use the browser and agree to have their web history tracked.
A move to a censorship model like the one Kuniao presents could solve problems that Chinese censors have been struggling with for decades: detecting traditional VPNs is computationally expensive and determining “good” (scientific research) from “bad” (reading NYTimes reporting on Xinjiang) firewall evasion is impossible.
Censorship now is largely overt. It is impossible to miss the fact that entire websites will fail to load at all.
However, through the Kuniao approach, censors can seamlessly edit out a handful of search results or tweets, even inserting fake articles into otherwise trustworthy websites.
An NYTimes article claiming that there is no mistreatment of Uighurs in Xinjiang is not out of the question, and if the censors are willing to invest enough resources into it determining fakes for the average user would be extremely difficult, akin to the difficult task of determining artwork forgeries. It also introduces the opportunity for developers to include general-purpose spyware like they did in the Study Xi Superpower app. In such a case, your microphones and webcams could be very easily compromised.
In order to effectively censor the traffic of China’s 854 million internet users, the government would need to look at every payment made with Alipay, every message sent on WeChat, every video watched on Youku, and anything else anyone does on the internet every hour of every day. Instead, censors choose to focus their resources where they need to and outsource what they can. Researchers have observed for almost a decade that government censorship happens at the border of the country, while inside the country website developers are expected to do it themselves. This is why a user in China will find that they cannot load Twitter at all, but searching for the Tiananmen Square Massacre on Weibo will simply load an empty search results page.
Censors also worry about blocking traffic that would be beneficial to other government initiatives. The ban on Google’s services, for example, creates a large problem for China’s researchers who rely on Google Scholar, and traditional methods of censorship are only becoming more and more ham-fisted.
Before the introduction of HTTPS, it was possible to simply monitor what is present on a webpage and block connections when the webpage displayed non-approved material. For instance, a Chinese user could browse Wikipedia, but they could not load the Falun Gong page. But the spread of HTTPS means that the content of webpages are encrypted, encrypts the content of webpages, making such surgical censorship impossible [For a while, the CCP compromised on Wikipedia, banning Mandarin-language articles but leaving English Wikipedia accessible]. Because of the introduction of HTTPS and its near-universal adoption outside of China, censors can no longer just partially ban websites. As a result of this restriction, websites like Github that are too important to ban have become place to post openly without fear of censorship.
The HTTPS shield only protects users while the data is encrypted, however. Ultimately, data is useless if it is never decrypted, and people rely on browsers for decrypting. Earlier this year in a move that foreshadowed the advent of Kuniao, Chinese browsers used this vulnerability to start censoring a GitHub page that contained protests against the 996 work schedule common at Chinese technology companies.
Moving censorship from the network to browser level means that censors have to pay the computational price of monitoring your traffic. Further, encryption is no longer a roadblock to targeted censorship because while you can hide from observers on the network, you can’t hide from your own browser. Just as censors previously told providers to censor on their behalf in the past to reduce the number of work censors had to do, they may soon tell users to censor themselves to reduce work and to render encryption useless.
Browsers such as Kuniao represent the possibility that in the future, censorship and surveillance will reach new levels of sophistication. Being able to use Western social media in a limited fashion is a spoonful of sugar to help the medicine go down.