On December 5, 2012, the US Congress unanimously passed a resolution with an extraordinary 397-0 vote calling for US opposition to government control of the Internet and the preservation of a multi-stakeholder model of Internet governance. The bipartisan resolution was responding to global proposals surrounding a 2012 Dubai gathering known as the World Conference on International Telecommunications (WCIT). The event was convened by the International Telecommunication Union (ITU), the UN specialized agency for informa- tion and communication technologies, to review the International Telecommunication Regulations (ITRs), a global treaty adopted in 1988 to establish guidelines for how telecommunications operators exchange traffic across national borders.

One of the more controversial proposals circulating in advance of the Dubai meeting, but which did not make it into the revised ITRs, emanated from an association of European telecommunication operators raising the prospect of “sending party network pays” in international interconnection. In traditional voice communication, the payment burden rests primarily with the originating caller and network. Extending this model to the Internet would potentially create an economic transformation amounting to a charge on content companies such as Google every time a user chooses to download a YouTube video. This type of regulatory and pricing burden on content providers would have democratic implications because it would potentially create economic disincentives for new content companies and innovations and also fragment the Internet if content companies opted to avoid these payments by withdrawing their content in parts of the world imposing fees. A separate proposal from Russia, China, and several Arab nations asserted the sovereign right of UN Member States to establish international policy in the realm of Internet governance.

Against a backdrop of hyperbolic media reports about the UN trying to “take over the Internet” and after two weeks of contentious deliberations, the conference ended with nations divided about whether they would sign a revised treaty. While the circumstances surrounding this epoch were technically and institutionally complex and shaped by a protracted history of global Internet governance tensions, the United States and many other advanced economies ultimately refused to sign the treaty because they objected to the inclusion of any language about the Internet, even as part of a non-binding resolution. These countries believed that the scope of the treaty, or even future discussions, should not be expanded to include Internet architecture or content but remain more narrowly tailored to telecommunications interconnection. The global controversy over this treaty provides a moment of opportunity to address the implications of Internet governance methods and structures for democracy.

Twenty-first century conflicts over Internet gover- nance are increasingly proxies for global political struggles. The same technologies creating new potentialities for citizen engagement and global access to knowledge are providing governments with extraordinary opportunities for censorship and surveillance. At the most extreme level, repressive nation states unable to influence the flow of citizen information by traditional means have turned to infrastructure control systems to terminate communication networks during political uprisings. For example, a prolonged late-2012 Internet blackout in Syria reportedly curtailed the ability of citizens to coordinate political action internally or disseminate information to the outside world via Internet videos or blogs about government atrocities.

Internet governance is enacted by private corporations and new global institutions as much as nation states. The corporatization of many aspects of Internet governance involves its own challenges related to the legitimacy of private entities establishing public policies related to individual civil liberties. Understanding connections between Internet governance and freedom requires understanding the underlying technological systems of Internet governance, the institutional frameworks of control over these systems, and the core policy decision points with heightened implications for the democratic public sphere. Internet governance decisions increasingly determine the extent of political and economic liberty. Multi-stakeholder governance of the Internet, though itself problematic, provides the distributed balance of power necessary to foster democratic conditions of Internet freedom.

The Internet and Political and Economic Struggle

A significant shift in the 21st century has been the co-opting of Internet governance technologies for political and economic uses quite distinct from their original design and function. One example of this phenomenon is the use of technologies of Internet governance to terminate access and services, such as the Egyptian Internet outage that lasted for several days when then-President Hosni Mubarak ordered a severing of communication services during political unrest directed against his regime. The Internet disruption prevented citizens from accessing the Web, sending email, or using social media.

One of many approaches for terminating Internet services is via the Internet’s routing and addressing infrastructure. Networks that interconnect to form the global Internet use a standard exterior gateway protocol called Border Gateway Protocol (BGP) to advertise to each other unique sets of Internet addresses that can be reached via each network. This protocol is a set of rules specifying how networks should broadcast and which customers, services, and web sites can be accessed within or through a network. As such, it is one of many core governing technologies keeping the Internet operational.

Unfortunately, this system can also be used to disrupt the Internet. Simply terminating these advertised routes makes everything that was previously accessible via a network disappear. All of the sites and resources are still present, but they are not reachable on the Internet. When the Egyptian Internet outage occurred, the collection of reachable Internet addresses advertised to the world from various Egyptian network operators simply disappeared. Cell phone service was also terminated, as the government ordered various network providers to suspend services.

The Internet’s Domain Name System (DNS) has also increasingly become a primary tool for blocking and filtering content. The DNS is a central technology of Internet governance because it translates between the alphanumeric names that humans use (e.g. www.lauradenardis.org) and the binary numbers that computers use to route information to its destination over the Internet. DNS filtering and redirection is one method China uses within its so-called Great Firewall. When an individual types a domain name, the domain name either fails to resolve into the requested site or is redirected to another site. This is technically easy to accomplish, usually involving a government official directing the Internet registry overseeing a particular top-level domain (e.g. web sites ending in “.cn”) to alter the record that translates a domain name into an Internet address.

The Domain Name System has also become a tool for content control economically related to intellectual property rights enforcement. For example, the Immigration and Customs Enforcement (ICE) agency of the US Department of Homeland Security directs Internet registries, such as .com registry VeriSign, to redirect domain names from sites that sell counterfeit goods or copyrighted media to a site with a law enforcement method.

When Wikipedia, Reddit, and other prominent Internet content companies blacked out their online sites in January 2012, this boycott was protesting legislation moving through the US Congress that would increase the use of the Domain Name System for copyright and trademark enforcement. The two proposed bills were the “Stop Online Piracy” Act (SOPA) and the “PROTECT IP” Act (PIPA). The boycotts were not protesting the anti-piracy intent of the bills but the mechanism for how intellectual property enforcement would be executed. Internet service providers would be asked to alter the authoritative records the Internet registries provided for resolving Internet names into addresses, possibly fragmenting the Internet and creating security concerns within the DNS. SOPA and PIPA were ultimately rejected.

What will be the consequences of this increasing interest in Internet governance technologies for controlling content online? Internet governance technologies have already been a politically-charged area of international relations, but that will only worsen as complex and topdown interventions increase. As Google’s Chief Internet Evangelist and the inventor of the core TCP/IP protocols on which the Internet operates, Vint Cerf, said to the US Congress in hearing testimony, “If all of us do not pay attention to what is going on, users worldwide will be at risk of losing the open and free Internet that has brought so much to so many.”

The Privatization of Governance

While government interventions in Internet gov- ernance have been controversial, privatized governance creates a different set of concerns. On September 11, 2012, an act of terrorism at the US diplomatic mission in Benghazi, Libya resulted in the killing of four US citizens including Ambassador Chris Stevens. The attack occurred contemporaneously with widespread protests over the release of an amateurish YouTube video entitled “The Innocence of Muslims.” During the height of rioting and protests across the Arab world, the Obama administration asked Google to remove the video from its YouTube platform. Google refused, indicating that the video failed to constitute a violation of its terms of service. Governments frequently ask private companies to remove content, whether for political reasons, law enforcement actions, or national security concerns. This is a very complicated area of content governance for corporations because they are subject to an enormous variety of laws varying from country-to-country in content regulation areas such as defamation, decency, and freedom of expression.

The informal corporate motto of Google is “Don’t be Evil.” The private corporations that serve as information intermediaries such as search engines, social media platforms, and content aggregation sites wield a tremendous amount of power over communicative expression and other core democratic freedoms. The adjudication of expressive freedom takes place in many ways, including discretionary and delegated censorship in which corporations determine whether to acquiesce to government requests to take down content or terminate a user account and through private decisions to remove content or terminate an online presence because of a violation of an end user agreement or some more arbitrary reason.

The terms of use policies and content removal deci- sions of companies, apart from government takedown requests, are privatized forms of governance that make decisions about privacy, reputation, cyberbullying, and free expression. During the 2012 London Olympics, Twitter suspended the personal Twitter account of a British jour- nalist. The company stated that it terminated the account at the National Broadcasting Corporation’s (NBC) request after the reporter tweeted the personal email of an NBC executive, a violation of Twitter rules. The journalist had been posting tweets criticizing NBC’s coverage of the games and, because Twitter had a cross-promotional partnership with NBC during the Olympics, it was criticized for terminating this account. Twitter ultimately restored the journalist’s account, but the fact that this termination was within Twitter’s rights as a private corporation sheds light on the democratic implications of the corporatization of Internet governance. Similar controversial examples of privatized and infrastructure-based governance of the public sphere followed the WikiLeaks release of US diplomatic cables. Online financial transaction companies, like PayPal, ceased providing services to WikiLeaks. Web hosting intermediary Amazon similarly cited a violation of its terms of service when it ceased providing web hosting services to WikiLeaks.

While these content and account intermediation decisions and delegated censorship requests are quite high-profile, privatized Internet governance exists in areas concealed in much more technological complexity. For example, much of the digital world is funded by online advertising. The data retention and aggregation practices of information intermediaries and third party online advertising companies influence the extent of individual privacy online. These types of privately-established and privately-mediated policies raise questions about the legitimacy of these organizations to set these policies, the role of corporate social responsibility, and the appropriate role of traditional governments as a check on this power. At a minimum, principles of disclosure and transparency of processes and some level of public accountability can provide some degree of consistency and accountability.

Limitations of Multi-Stakeholder Governance

The Internet is already governed, just not in the traditional hierarchical sense of sovereign nation-state or intergovernmental governance. From a technical and administrative standpoint, Internet governance involves the design and administration of technical architecture and the enactment of substantive public policy issues at technical control points. This technical Internet governance is enacted through the design of technical architecture, the privacy policies of private corporations, the coordination of critical Internet resources by relatively new but increasingly powerful global institutions, the agreements of international treaties, and the laws of traditional governments within national borders.

For example, technical standard-setting institutions such as the Internet Engineering Task Force (IETF), the World Wide Web Consortium (W3C), and the Institute for Electronics and Electrical Engineers (IEEE), design the standards (e.g. HTTP, XML, Wi-Fi) necessary for interoperability among devices. These standards perform a highly technical function but also make various public policy decisions. The functions and critical Internet resources (CIRs) that ICANN (Internet Corporation for Assigned Names and Numbers) oversees are another form of technical governance. Unique CIRs such as domain names and Internet addresses are necessary to keep the Internet operational and the administration and control of these finite resources has historically been a contentious topic, particularly because of a global impasse over the United States government’s historic connection to ICANN and the new power of the Regional Internet Registries (RIRs) in controlling increasingly scarce Internet addresses.

A constantly shifting balance of powers between private industry, international technical governance institutions, governments, and civil society is necessary to create democratic conditions for online economic and expressive liberty. This balance of powers is often called “multi-stakeholderism.” The mantra of multi-stakeholder governance has been elevated to a value in itself in the Internet governance arena, primarily because it serves as a foil against the prospect of greater government restrictions on political and economic freedom. Unfortunately, the concept of multi-stakeholderism can also be deployed as an impediment to the diverse types of governance, or absence of governance, necessary for promoting conditions of democracy.

The limitations of multi-stakeholderism arise in several forms, including the democratic challenges presented by private sector governance and the risk that multistakeholderism becomes a formalized mechanism of multi-state power rather than power distributed more broadly among civil society, private industry, global institutions, and formal nation states. Multi-stakeholder approaches in international arenas can become centralized governance processes because they, to be multi-stakeholder in practice, require some type of formal procedural mechanism, and even hierarchy, to ensure that appropriate voices are heard. The argument of some is that only governments have the necessary legitimacy to create and enforce processes that are adequately multi-stakeholder. The top-down procedural enforcement this entails is exactly the opposite of the Internet’s bottom-up and expertise-based administration, such as traditional standards-development procedures of the Internet Engineering Task Force.

Enforcing democratic conceptions of multi-stakeholderism in international fora similarly raises the question of whose conception of democracy to instantiate in Internet policies. For example, China and Russia are prominently involved in multi-stakeholder discussions in United Nations fora. Their conceptions of implementing democratic values on the Internet are more concerned with restrictions on access to knowledge than with freedom of expression. Finally, multi-stakeholder deliberations, such as the World Conference on International Telecommunications mentioned earlier, often become relegated to multiple governments rather than multiple stakeholders, with non-governmental actors limited from participating in formal deliberations and lacking any voting power.

A formalized system of intergovernmental control over Internet governance could produce any number of adverse consequences such as limiting private industry and civil society input into policy decisions, centralizing an inherently decentralized system, slowing innovation and growth, and creating additional administrative pathways for surveillance and censorship. As governments increasingly turn to Internet governance technologies like protocols and the Domain Name System for content control—whether good or bad—it is not surprising that there is increasing interest in gaining more formal jurisdiction for controlling these infrastructures.

Despite the limitations of multi-stakeholder Internet governance, it has succeeded in keeping the Internet operational and providing possibilities for access to knowledge and free expression. The next tests for the efficacy of multi-stakeholder governance will occur in addressing emerging challenges to democratic Internet freedom, such as the increasing privacy concerns over the practices of online advertising, government proposals to limit personal anonymity online, and the trend away from universal Internet interoperability in favor of proprietary approaches that privilege dominant corporations. On these and other policy issues surrounding the online public sphere, it is vital that the public is engaged in these debates, because as goes Internet governance so goes Internet freedom.