How to Free the Chinese Internet

The Chinese Internet is becoming a walled garden -- an Orwellian environment, separated from the rest of the global Web, where information unfavorable to the government simply disappears, public discussions are shaped by undercover agents, and censorship and surveillance are built into the most popular online services.

This system is bad for the United States, not just because it contradicts American values and holds back China’s political liberalization but also because it harms U.S. businesses. Part of the problem is that U.S.-based online services, such as Facebook and Twitter, are blocked from reaching Chinese consumers, but the true stakes are much larger: Every American business that operates in China needs a secure, fast, and reliable Internet connection. And China’s censorship system puts these connections at risk.

The core of China’s repressive online system is its so-called Great Firewall, which makes large sections of the Internet unreachable from the Chinese mainland and steers Chinese users toward censored domestic sites. As part of its worldwide “Internet freedom” agenda, the United States has funded the creation of special software tools to help Chinese dissidents, journalists, and other users for whom privacy is paramount sneak through their government’s barriers. This focus on human rights, high-risk users, and customized software has worked well in some countries, but in China, U.S. policymakers need to take a broader view.

The existing U.S.-supported tools aim to give high risk users the protection they need -- protection far beyond what mainstream computing tools provide. But such cloak and dagger methods tend to be too slow and unreliable for most users. And each time these tools improve, Beijing’s army of censors adapts in response. China has bottomless reserves of high-caliber engineering talent to devote to shoring up its barricades, and it has developed advanced techniques to identify the distinctive signature of particular censorship circumvention programs, blocking them automatically or leaving them barely usable. Tools built explicitly to support free expression, which lack a broad commercial footprint, are ultimately no match for this pressure.
 
At the same time, the U.S. focus on China’s digital dissidents neglects a much larger group of Chinese Internet users. Most Chinese citizens want Internet access for personal, not political, reasons -- think of young workers in the technology industry and students in Chinese universities. They want to watch popular videos more than to read anti-regime material, to tweet more than to plan protests. They do not expect or need their activities to be clandestine -- they just want a fast and unrestricted connection. Chinese students try to circumvent censorship in roughly the same way that American students share pirated movies: it’s against the rules, and the authorities warn against it, but by itself it seldom gets people into serious trouble.

In order to help the Internet unlock political change in China, American policymakers should focus on defending the commercial needs of U.S. businesses operating in China, rather than the political needs of dissidents and journalists. A Chinese Internet that allowed free and private online communication would be valuable not just to Chinese users (dissident and otherwise) but also to the many American businesses whose offices in China need to communicate with their U.S.-based headquarters. In other words, making the Chinese Internet safe for American business might be the best way to advance human rights online.

The need for a shift in U.S. Internet freedom policy toward China was made clear by a recent survey of more than a thousand Chinese users of circumvention software, conducted by my colleagues and me at the Open Internet Tools Project. We found that where politically motivated software is failing to give Chinese users secure and free Internet access, business tools are succeeding. Every foreign firm that operates in China needs a secure, fast, reliable, and uncensored connection to its home office. Tens of thousands of such connections break through the Great Firewall every day, including Virtual Private Network (VPN) connections and encrypted connections to cloud-hosted business services (such as secure websites for holding corporate documents). These types of connections, rather than the niche circumvention tools that currently receive the most U.S. support, are increasingly offering Chinese users their best option for Internet freedom.
 
Chinese authorities can see when a user connects to the Internet through VPNs or cloud-hosted servers, but not which specific sites he visits. They could decide at any moment to block all such connections and only allow through the traffic that they can directly monitor. But they have chosen not to take this step, since the collateral economic cost of blocking all this encrypted traffic (much of which is corporate) would simply be too high.
 
Instead, China has charted a middle course, intermittently degrading the connections but generally letting them through. That does not mean that the connections are perfect: U.S. firms suffer significant costs and frustration on account of the firewall. In a recent survey of American executives working in China, over half reported that Internet censorship hurt their company’s ability to conduct business there. But these businesses still enjoy much better Internet access than Chinese citizens themselves.
 
More and more, Chinese users are taking advantage of this trend, using platforms primarily designed for business to camouflage their activity behind commercially valuable traffic. In total, 64 percent of the respondents to our survey rely on such platforms. The single most widely used tool, GoAgent, uses a business-oriented cloud system provided by Google called App Engine. App Engine was built for software developers, but GoAgent lets any Chinese Internet user harness it as a personal proxy server, allowing him to reach censored websites, email servers, or social media services.
 
The Chinese government, in turn, is trying to better segregate consumer from corporate Internet traffic. Late last year, for example, it passed a new law requiring that providers of “website access services” (including VPNs) collect the real name of each user who registers. This would make it easier for the authorities to crack down on Chinese citizens who use VPN technology, while still allowing foreign businesses to use it. Chinese officials have also pressured some foreign businesses to install on their computers special monitoring hardware controlled by the Chinese police, which might let the authorities monitor American businesses’ internal communications and Internet use. (The U.S. embassy is reportedly working behind the scenes to help American companies resist any such mandate.)
 
The ways in which Chinese users get around barriers to the Internet carry clear lessons for U.S. policy. The United States should shift its focus and approach Chinese censorship less as a human rights issue than as a matter of commercial diplomacy. If Washington can defend American businesses’ legitimate interest in fast, reliable, and secure connections, it will end up bestowing those same connections on many Chinese users.
 
To begin with, Washington should reframe its protests with Beijing about Internet freedom. China surely knows that its economic growth depends on creating a welcoming environment for foreign firms; the United States should remind it that ensuring reliable Internet access is a crucial part of doing so. Moreover, U.S. officials should change the way they talk about the trade implications of Internet freedom. Washington has argued that Chinese censorship functions as an illegitimate trade barrier inconsistent with China’s WTO commitment to free trade in services, since the firewall prevents U.S.-based Internet companies such as Google from offering high quality service directly to Chinese consumers. The U.S. trade representative formally raised this issue with the Chinese in 2011, but WTO rules, which make “protection of public morals” a valid basis for trade barriers, are unlikely to vindicate this U.S. argument. On the other hand, the firewall’s disruption of secure corporate communications between the United States and mainland China does nothing to protect public morals, and may be a less defensible practice under WTO rules.
 
The second element of a new, commerce-focused approach to Internet freedom should be to place special emphasis on the information needs of smaller U.S. businesses operating in China. A Fortune 500 company with an office in Shanghai might invest in its own physical wireline connection all the way to Hong Kong, but an American small-business owner traveling to visit a Chinese factory will more likely rely on a standard mainland Internet connection, and will need to reach cloud-based information services. Such services are also well positioned to meet the needs of individual Chinese users.
 
If the United States can assure that fast, reliable, and secure communications make it through the Great Firewall, then Chinese users themselves will find clever ways of piggybacking on that traffic. It is telling that GoAgent -- one of the most popular circumvention tools in China today -- was written by Chinese programmers.

Ultimately, a freer Internet would bring China closer to the rest of the world and create more favorable conditions for political reforms. The existing U.S. Internet freedom programs are vital for dissidents and vulnerable users in many repressive regimes, and they should be continued. But for China, their aims would be better achieved with a different, business-focused approach.