Under conditions like these, the world will see its first Internet asylum seeker. A dissident who can’t live freely under an autocratic Internet and is refused access to other states’ Internets will choose to seek physical asylum in another country to gain virtual freedom on its Internet. There could be a form of interim virtual asylum, where the host country would share sophisticated proxy and circumvention tools that would allow the dissident to connect outside. Being granted virtual asylum could be a significant first step toward physical asylum, a sign of trust without the full commitment. Virtual asylum would serve as an extra layer of vetting before the physical asylum case reached the courts.
Virtual asylum will not work, however, if the ultimate escalation occurs: the creation of an alternative domain name system (DNS), or even aggressive and ubiquitous tampering with it to advance state interests. Today, the Internet as we know it uses the DNS to match computers and devices to relevant data sources, translating IP addresses (numbers) into readable names and vice versa. The robustness of the Internet depends on all computers and networks’ using the same official DNS root (run by the Internet Corporation for Assigned Names and Numbers, or ICANN), which contains all the top-level domains that appear as suffixes on web addresses—.edu, .com, .net and others.
But there are alternative DNS roots in existence, operating in parallel with the Internet but not attached to it. Within tech circles, most believe that the creation of an alternative DNS would go against everything the Internet represents and was built to do: namely, share information freely. No government has yet achieved an alternative system,3 but if a government succeeded in doing so, it would effectively unplug its population from the global Internet and instead offer only a closed, national intranet. In technical terms, this would entail creating a censored gateway between a given country and the rest of the world, so that a human proxy could facilitate external data transmissions when absolutely necessary—for matters involving state resources, for instance.
For the population, popular proxy measures like VPNs and Tor would no longer have any effect because there wouldn’t be anything to connect to. It’s the most extreme version of what technologists call a walled garden. On the Internet, a walled garden refers to a browsing environment that controls a user’s access to information and services online. (This concept is not limited to discussions of censorship; it has deep roots in the history of Internet technology: AOL and CompuServe, Internet giants for a time, both started as walled gardens.) For the full effect of disconnection, the government would also instruct the routers to fail to advertise the IP addresses of websites—unlike DNS names, IP addresses are immutably tied to the sites themselves—which would have the effect of putting those websites on a very distant island, utterly unreachable. Whatever content existed on this national network would circulate only internally, trapped like a cluster of bubbles in a computer screen saver, and any attempts to reach users on this network from the outside would meet a hard stop. With the flip of a switch, an entire country would simply disappear from the Internet.
This is not as crazy as it sounds. It was first reported in 2011 that the Iranian government’s plan to build a “halal Internet” was under way, and more than a year later it seemed that the official launch was imminent. The regime’s December 2012 launch of Mehr, its own version of YouTube with “government-approved videos,” added yet another data point that the regime was serious about the project. Details of the plan remained hazy, but according to Iranian government officials, in the first phase the national “clean” Internet would exist in tandem with the global Internet for Iranians (heavily censored as it is), then it would come to replace the global Internet altogether. This would entail moving all the “halal” websites to a particular block of IP addresses, which would make it trivially easy to filter out websites that are outside the halal block. The government and affiliated institutions would provide the content for the national intranet, either gathering it from the global web and scrubbing it, or creating it manually. All activity on the network would be closely monitored, facilitated by the government’s top-level infrastructure control and agency over software (something Iranian officials are very concerned about, judging from a 2012 ban on the import of foreign computer security software). Iran’s head of economic affairs told the country’s state-run news agency that they hoped their halal Internet would come to replace the web in other Muslim countries, too—at least those with Farsi speakers. Pakistan has pledged to build something similar.
It is possible that Iran’s threat is merely a hoax. How exactly the state intends to proceed with this project is unclear both technically and politically. How would it avoid enraging the sizable chunk of its population that has access to the Internet? Some believe it would be impossible to fully disconnect Iran from the global Internet because of its broad economic reliance on external connections. Others speculate that, if it wasn’t able to build an alternative root system, Iran could pioneer a dual-Internet model that other repressive states would want to follow. Whichever route Iran chooses, if it is successful in this endeavor, its halal Internet would surpass the Great Firewall of China as the single most extreme version of information censorship in history. It would change the Internet as we know it.
Virtual Multilateralism
In parallel with these balkanization efforts, we will see the rise of virtual multilateralism based on ideological or political solidarity, involving both states and corporations working together in official alliances. States like Belarus, Eritrea, Zimbabwe and North Korea—authoritarian, with strong personality cults and a pariah status elsewhere in the world—would have little to lose by joining an autocratic cyber union, where censorship and monitoring strategies and technologies could be shared. As these countries collaborated to build virtual-age police states, it would become increasingly difficult for Western companies, from a public-relations standpoint, to conduct business there, even if it was legal. This would create space for non-Western companies, whose shareholders may have fewer qualms and who are used to working in similar environments, to play a more active business role within a network of autocratic states.
It’s no accident, for example, that the company that owns 75 percent of North Korea’s only official mobile network, Koryolink, is the Egyptian telecom Orascom, a firm that thrived under the long reign of Hosni Mubarak. (The other 25 percent is owned by North Korea’s Ministry of Posts and Telecommunications.) For North Korean subscribers, Koryolink service is a walled garden, a highly limited platform that allows for only basic functionality. Koryolink users can’t make or receive international calls; nor can they access the Internet. (Some people can access the North Korean intranet, an odd pastiche of online content, mostly propaganda, that government officials transfer over from the Internet.) Local phone calls and text messages are almost certainly monitored, and The Economist reported that the network is already a platform for the dissemination of government propaganda, with the North Korean daily Rodong Sinmun sending users the latest news by text message. While it is not officially a requirement, most people are “encouraged” to pay their phone bills in euros (which are unofficially in circulation), a tall order for most North Koreans. Even so, the demand for phones was so great that adoption soared in the country, leaping from three hundred thousand subscribers to more than a million within an eighteen-month period ending in early 2012. Koryolink’s gross operating margin of 80 percent means big business for Orascom.