Jump to ContentJump to Main Navigation
Negotiating Internet Governance$

Roxana Radu

Print publication date: 2019

Print ISBN-13: 9780198833079

Published to Oxford Scholarship Online: April 2019

DOI: 10.1093/oso/9780198833079.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 06 December 2021

The WSIS Decade and the Public–Private Partnership Thirst

The WSIS Decade and the Public–Private Partnership Thirst

(p.113) 5 The WSIS Decade and the Public–Private Partnership Thirst
Negotiating Internet Governance

Roxana Radu

Oxford University Press

Abstract and Keywords

The governance of the Internet faced a reflexive turn throughout the WSIS decade (2005–15), explored in this chapter. Concerns for authority, legitimacy, and accountability—expressed by different stakeholders—became central to the evolution of the field. A number of challenges, stemming from three diverse sources, were embedded therein. First, questions were spawn by the modus operandi of the sui generis institutions, such as the international technical bodies exercising public governance functions to ensure the continuous functioning of the Internet. Second, demands resulted from the gradual adaptation of intergovernmental organizations with core or tangential interests in the field. Third, the role of private intermediaries was called into question as their financial and political power rose steeply. Their relation to governments was also probed, particularly after the 2013 Snowden leaks.

Keywords:   internet governance, WSIS, WGIG, IGF, WCIT, Snowden, stakeholders, civil society, negotiation

When the United Nations convened its first World Summit on the Information Society (WSIS) in Geneva in 2003, what surfaced was the merging of the ICT agenda with concerns regarding the uneven expansion of the Internet and imbalanced representation of states in its governance mechanisms. The Summit, split in two phases (Geneva 2003 and Tunis 2005), was the starting point for vocalizing the worries of developing countries, but also for collaborative institutional endeavours that would continue throughout the decade in what became known as the ‘WSIS process’. The sheer breadth of topics covered by this process and the formal recognition of shared responsibility of different actors marked a turning point in the evolution of the Internet, setting the stage for an accelerated institutionalization of an emerging policy field.

Sprouting from the WSIS process, the tensions regarding the role of the United Nations and that of the United States in Internet-related decision-making were particularly significant for the evolution of Internet governance (IG). The WSIS Summit Outcome document contributed to the formation of IG as a public domain, providing a conceptual framework for understanding it and delimiting mandates and responsibilities. By sanctioning a broad definition for IG, WSIS provided a common vocabulary shaping the conversations for years to come. Mirroring the Geneva Declaration of Principles adopted two years before, the Tunis phase recognized the need to continue to involve a variety of stakeholders in relevant governance arrangements, without going much further in clarifying responsibilities for this hybrid order.

The period comprised between 2005 and 2015 also shed light on the regulatory efforts of various actors, primarily states and international organizations. The rise of smartphones and mobile Internet was unparalleled: the global mobile data traffic grew 4,000-fold between 2005 and 2015 (Cisco 2017) and facilitated the fastest expansion of connectivity worldwide. Accordingly, the governance arrangements of the time revealed a stronger involvement of corporations and civil society in IG processes, alongside a redesign of the (p.114) public–private sector relationships, tested in watershed moments such as the 2012 World Conference on International Telecommunications or the 2013 Snowden leaks on mass surveillance. They also revealed stronger collaboration between governments and private actors on content controls in the context of the Arab Spring revolutions between 2010 and 2012, followed by the introduction of policing regimes for online activities. This chapter traces the evolution of the field throughout the WSIS decade, providing an analysis of the governance transformations that marked the field’s coming of age.

Internet Governance @ WSIS

In both academic and policy discussions, WSIS is known to have given the first definition of IG endorsed in a globally negotiated outcome document. With its focus on a broadly understood ‘information society’, WSIS constituted a UN summit comparable in scope and purpose to the 1992 Earth Summit, the 1995 World Conference on Women, and the 2002 World Summit on Sustainable Development. The idea for a large-scale event that would tackle ICTs and development came from the International Telecommunication Union (ITU) in 1998 and was authorized through General Assembly Resolution 56/183 in 2001. The Summit was to link with the UN Millennium Declaration, starting a discussion around the use of ICTs to achieve the Millennium Development Goals (MDGs).1 The Summit was held in two phases: the first took place in Geneva from 10 to 12 December 2003 and the second was held in Tunis from 16 to 18 November 2005; the preparations for and follow-up to WSIS-related activities are referred to as the ‘WSIS process’.

The two Summit phases were preceded by PrepCom meetings, consultations, working group sessions, and roundtables which included a range of stakeholders, from UN member states to civil society actors. The process of designing the summit benefited from input gathered in three global preparatory meetings held in Geneva between July 2002 and September 2003, and four regional meetings in Africa, Asia-Pacific, Europe, and Latin America and the Caribbean. The process was open to non-state actors, who were allotted the status of observers. Hitherto, practices of inclusion in working groups varied throughout the preparations, from government delegates only to equal footing partaking by different sector representatives. In December (p.115) 2003, more than 11,000 participants attended WSIS-I in Geneva. Four preparatory global meetings (all in Geneva) and three regional gatherings took place before WSIS-II and over 19,000 representatives were present in Tunis in November 2005. The significant growth in attendance for WSIS-II was mainly due to a two-fold increase in the number of civil society representatives and a ten-fold increase in business sector participation. In total, the Summit attracted nearly fifty heads of state/government and vice-presidents and twice as many ministers and vice-ministers, the majority from developing countries.

Against the backdrop of state-negotiated telecommunication regimes, WSIS represented an unprecedented experiment, with 45 per cent of all participants in the Geneva phase and 60 per cent in the Tunis phase being non-state actors (Chenou 2014). Apart from the ‘usual suspects’ for large-scale UN summits, most of the attendees reflected the variety of entities that were directly affected by digital convergence. This phenomenon challenged the works of different sectors (telephony, radio, recorded music, film, video, books, magazines, and libraries), affecting the way in which they were governed by separate policies before. The governance of ICTs and media was traditionally placed under the control of states (Irion and Radu 2013) and the publishing, film, and music industries maintained strong positions on intellectual property rights. Their presence at the Summit meant that these issues would be addressed in the negotiations. Notably, technical Internet bodies such as the Internet Corporation for Assigned Names and Numbers (ICANN) or the Internet Engineering Task Force (IETF)/Internet Society (ISOC) were well-represented in the consultations. Last but not least, the business sector secured a double representation,2 first as corporate entities speaking on their own behalf and second via the Coordinating Committee of Business Interlocutors (CCBI) organised by the International Chamber of Commerce (Raboy et al. 2010).

The WSIS process placed a strong emphasis on the developmental dimension of the information society, with a view to addressing the perceived inequalities. On the one hand, it discussed the terms of access to ICTs, in particular stressing the availability, quality, cost, and capability to connect to the Internet in developing countries. In this regard, WSIS called into question the commercial arrangements for traffic services established in the 1990s, resulting in a net flow of revenues into developed countries and higher costs for access in the global South. Moreover, capacity building was deemed necessary (p.116) for developing the skills required to thrive in the information society. On the other hand, in addressing governance mechanisms, the focus was on the representation of developing countries in global ICT-related arrangements. Salient in the WSIS negotiations were two issues: first, gaining access to processes by which the Internet was managed; second, ensuring an equal say in debates that have been transformed by the Internet, such as intellectual property, trade, or consumer protection.

Without mentioning IG, the January 2002 UNGA resolution 56/183 announcing the Summit set in motion a process that would redefine the understanding of the terms, but also the rules of engagement. Participants (Drake 2005; Kummer 2007) recall that IG was not among the topics originally discussed and was added to the WSIS agenda gradually. Among the factors that increased the visibility of this topic were the intense IG discussions held throughout the ITU plenipotentiary conference of Marrakesh in September to October 2002 (marking a divide between the United States and developing countries on the management of the domain name system (DNS)) and the development of internationalized and multilingual domain names. IG was thus closely related to unresolved technical issues plus institutional turf battles and resurfaced as a subset of ICT for development (ICT4D) concerns.

In Search of a Definition

While the ITU assumed the leading managerial role, the Summit was facilitated by an open-ended intergovernmental Preparatory Committee (PrepCom) defining the agenda and the means of participation, as well as drafting the outcome documents of the Summit. Fewer than 1,000 people participated in the Geneva preparations for WSIS-I, which constituted an important arena for negotiation. The first PrepCom meeting on 1 to 5 July 2002 in Geneva was dominated by representatives of states and civil society.3 The elected President of the PrepCom was Adama Samassekou (Mali), with elected Vice-Presidents from five geographic groups: Africa, Asia, Eastern Europe, Western Europe, and Latin America and the Caribbean, in addition to the two co-hosts (Switzerland and Tunisia) added ex officio.

(p.117) The themes put forward at the first PrepCom for initial discussions included access to ICTs, development of a policy and regulatory framework, ICT applications, infrastructure and network security, capacity building, as well as the role of different stakeholders in the promotion of ICTs for development. Calls to discuss IG were made by the European Union,4 Brazil,5 and a coalition of twenty-two NGOs, the latter calling, among other, for the democratization of ICANN. Notably, adding IG to the agenda meant the Internet was understood as a subcategory of the governance of the information society, with a particular set of concerns, the most poignant of these being its private nature. Integrating IG into the global governance approach shifted the focus away from the Internet exceptionalism discourse dominant in the 1990s (Chenou 2014).

Input from regional preparatory meetings and thematic multi-stakeholder roundtables on the discussion themes was presented at the second PrepCom meeting on 17 to 28 February 2003, where drafts for the Action Plan and WSIS declaration were also introduced. They were the product of a working group originally opened to all states and subsequently also open to all observers, chaired by Lyndall Shope-Mafole from South Africa. By the time the PrepCom for WSIS met for the third time in Geneva (PrepCom-3) on 15 to 26 September 2003, the draft text of the Declaration of Principles had already stirred controversy in the intersessional meeting on 18 July in Paris, where an Internet Governance Ad Hoc Working Group was created to discuss the issue. It was composed of state representatives and remained closed to non-state actors (with the exception of the first meeting). The proposed text reported at PrepCom-3 evidenced dissent around the formulation of roles and responsibilities outlined in paragraph 44 of the Declaration of Principles draft.

While they all agreed on the democratic, multilateral, and transparent management of the Internet, the opinions consolidated during the negotiations diverged in terms of participation entitlements, coverage, and role of states. I include below four alternative formulations that illustrate the divergent (p.118) perspectives of key stakeholders in the WSIS preparations, as they appear in the final report presented by PrepCom-3 in September 2003.

Formulation 1

proposed by the drafting group on Internet management

The international management of the Internet should be democratic, multilateral, transparent and participative with the full involvement of the governments, international organisations, private sector and civil society. This management should encompass both technical and policy issues. While recognizing that the private sector has an important role in the development of the Internet at the technical level, and will continue to take a lead role, the fast development of Internet as the basis of information society requires that governments take a lead role, in partnership with all other stakeholders, in developing and coordinating policies of the public interests related to stability, security, competition, freedom of use, protection of individual rights and privacy, sovereignty, and equal access for all, among all the other aspects, through appropriate [intergovernmental/international] organisations.

Formulation 2

Original text from 21 March document, supported by Saudi Arabia

Internet governance must be multilateral, democratic and transparent, taking into account the needs of the public and private sectors as well as those of the civil society, and respecting multilingualism. The coordination responsibility for root servers, domain names, and Internet Protocol (IP) address assignment should rest with a suitable international, inter-governmental organization. The policy authority for country code top-level-domain names (cc-TLDs) should be the sovereign right of countries.

Formulation 3

proposed by the EU

The international management of the Internet should be democratic, multilateral and transparent. It should secure a fair distribution of resources, facilitate access for all and ensure a stable and secure functioning of the Internet. It should respect geographical diversity and ensure representativeness through the participation of all interested States, including public authorities with competence in this field, of civil society and the private sector, with due respect to their legitimate interests.

Final text, in paragraph 48 of the Geneva Declaration of Principles

The Internet has evolved into a global facility available to the public and its governance should constitute a core issue of the Information Society agenda. The international management of the Internet should be multilateral, transparent and democratic, with the full involvement of governments, the private sector, civil society and international organizations. It should ensure an equitable distribution of resources, facilitate access for all and ensure a stable and secure functioning of the Internet, taking into account multilingualism.

The two basic visions for IG—intergovernmental control or private sector leadership—remained the Gordian knot of debates throughout the WSIS process. The wording finally adopted at WSIS-II was strongly normative, yet remained ambiguous as to the specific responsibilities or lead roles. At the WSIS-I meeting in Geneva, a Declaration of Principles was adopted, as (p.119) well as a Plan of Action aiming to provide guidance on the implementation of the first. As in previous negotiations, most contention emerged around the definition of IG, which would have elucidated the boundaries of participation and the extent to which the status quo was supported, as well as the roles of new actors in IG. The multi-stakeholder nature of the information society was not homogeneously understood by the delegations present at WSIS. What surfaced during the negotiations, before and after phase one in Geneva, was a split conceptualization of what multi-stakeholder governance meant: sometimes, the process was legitimized as a balanced representation of the state next to other stakeholders, via an equal number of members; other times, the state was understood as having a leading role in multi-stakeholder governance.

Moreover, strong contestation came from civil society groups which opposed the very notion of an ‘information society’ and, concordantly, opposed the organization and the Summit as capitalist endeavours (Raboy et al. 2010, 106–7). Hacker collectives refused to engage with the summit, while artists and activists designed alternative events, such as ‘WE SEIZE!’, which took place in front of the Palexpo complex where the Geneva Summit was held. The Geneva police intervened to disperse participants at other alternative events taking place throughout the city.

To overcome the lack of consensus discernible at the Geneva Summit, Kofi Annan, then Secretary-General of the United Nations, requested the establishment of a Working Group on Internet Governance (WGIG). The formation of this group was the only specific reference to IG in the Geneva Plan of Action. This WSIS outcome document translated the vision into eleven concrete action lines, with tasks mandated for governments including: facilitating the establishment of national and regional Internet Exchange Centres; managing or supervising, as appropriate, their respective country-code top-level domains (cc-TLDs); promoting, in cooperation with the relevant stakeholders, regional root servers and the use of internationalized domain names in order to overcome barriers to access, user education, and awareness about online privacy and the means of protecting privacy (2003a). An earlier reference to the contentious ‘governments should work to internationalize the management of Internet resources in order to achieve a universally representative solution’ (Draft Plan of Action, 21 September 2003, para. 19) was removed from the final Plan of Action. The private sector lead and the calls for an enabling environment had been toned down to:

Governments, in collaboration with stakeholders, are encouraged to formulate conducive ICT policies that foster entrepreneurship, innovation and investment, and with particular reference to the promotion of participation by women. (WSIS 2003a)

(p.120) The role of governments in the international governance of ICTs, including, inter alia, the naming and addressing resources, continued to generate debate. Many discussions called into question, implicitly or explicitly, the special role of one government: the United States, whose approach was challenged in three respects. First, authorizing the publication of any modifications, deletions, or additions to the root zone file, via NTIA, was seen as an imminent threat. Some countries perceived this as a ‘privilege’ that would give the United States the power to remove specific cc-TLDs from the root and eventually restrict the communication of users under that cc-TLD. Second, the location on American soil of ten out of the thirteen root servers6 was challenged. The third source of contestation was the informality of the arrangements among root server operators (Kleinwaechter 2005). The debate on IG at WSIS-I remained polarized, resulting in a compromise text focused on process rather than on substance. The proposal to establish a WGIG, mandated to develop a working definition of IG, was welcome. The expert group was also tasked to identify the public policy issues that were relevant to IG and develop a common understanding of the respective roles and responsibilities of governments, the private sector, and civil society.

Established in November 2004,7 the WGIG was chaired by Nitin Desai, the Special Adviser to the UN Secretary-General for the WSIS, appointed directly by Kofi Annan. The group consisted of forty experts equally representing the tripartite classification of stakeholder groups (states–businesses–civil society),8 but participating in an individual capacity. The majority of the selected members took part in WSIS-I and in the work of the UNICTTF or the G8 DOT Force. They had complementary skills and knowledge, ranging from technical specializations to development policy. A former Swiss diplomat, Markus Kummer was named executive coordinator of the working group after Switzerland announced that it would fund a secretariat for this new body.9 The Secretariat was established in July 2004, and on 20 to 21 September that year it held a two-day open consultation on the composition of the WGIG and its agenda, to which multiple stakeholders were invited. For the selection (p.121) of the WGIG members, broad criteria were put forward by Desai: regional representation, stakeholders, gender, developed and developing countries, and differing schools of thought; yet the final selection of participants was made by the Secretariat using informal means of consultation (Mathiason 2009). The largest part of group members came from governments (44 per cent), with a significant representation of emerging economies (including Brazil, China, Cuba, Iran, Russia) alongside European states. About a quarter of the WGIG members were active in ICANN discussions (Chenou 2014).

The WGIG Process

The WGIG prompted a momentous political negotiation and represented an early example of a new anchoring practice in use throughout the WSIS decade: the reliance on expertise from senior members of the community to find a solution to a stalemate. The group was at the epicentre of a much larger IG argument, marked by failure to agree on a shared understanding of the field. At least six international meetings within and outside the UN system touched upon this issue earlier that year, either in preparation for the second phases of WSIS or in specialized conferences. All ICANN meetings from July 2004 onwards also held dedicated sessions. To advance the WSIS negotiations, the task of scoping IG was allocated to WGIG in 2004. This challenge was addressed procedurally by designing a process that consisted of four multi-stakeholder meetings, all including an open consultation part, deemed necessary to ‘meet the concerns of those countries that did not want a small group process, but rather a full intergovernmental meeting’ (Desai 2005, ix). Four open consultations with stakeholders took place at the UN headquarters in Geneva between November 2004 and June 2005; representatives of states, business, and civil society actors were heard during the consultations in plenary session. Beside these, the WGIG members also met in closed sessions.

The WGIG mandate was subject to repeated contest, in particular for clarifying whether a reform of the IG system was at stake. For some, the work of the group was about the narrow technical side of the Internet. At the first WGIG meeting, ITU Secretary General Utsumi recommended ‘focus[ing] on the core activity of the management of Internet resources by ICANN, in particular top-level domains, which is where important issues remain unresolved’. For others, the WGIG task was inherently political.10 Former WGIG member William Drake summarized this as follows:


If I had a one US dollar for every time I was told—almost exclusively by my fellow Americans—that the WGIG was actually a UN plot to ‘take over the Internet’ and give it to the International Telecommunication Union (ITU), I could have a quite nice dinner—in Geneva no less. If in addition I had one US dollar for every time I was told that the WGIG was a plot against the ITU, I could have a nice dessert, and coffee too. (2005, 249)

The process that led to the final report of the WGIG was a collaborative exercise. A strong leadership role was taken by the Secretariat, led by Switzerland’s Markus Kummer. Following four open consultations, the WGIG published its report in July 2005 in preparation for the November meeting in Tunis. The Final Report was drafted in seventy-two hours at the Château de Bossey retreat outside Geneva (McLean 2005). A decision was taken by the WGIG together with the WSIS Secretariat to have a short final report transmitted to the UN Secretary-General for WSIS-II and a longer version as a Background Report. In its report, the group gave more leverage to public policy11 than to the technical standardization or technical resource allocation and assignment, thus enlarging the space for institutional engagement. In addition to Internet stability, security, and cybercrime, it covered intellectual property rights and trade-related issues, freedom of expression, data protection, and multilingualism. But it also included issues such as meaningful participation in global policy development, capacity building, or consumer rights, reflecting, to a large extent, the contribution of civil society participants.

The working definition it provided for IG sought to balance descriptive and normative dimensions and be as comprehensive as possible. Reflecting the academic influence in the group, WGIG equated IG with ‘the development and application by Governments, the private sector and civil society, in their respective roles, of shared principles, norms, rules, decision-making procedures, and programmes that shape the evolution and use of the Internet’ (WGIG 2005). Inspired by Krasner’s 1983 international regimes definition, this wording garnered the support of the working group members and, subsequently, of the WSIS delegations, which agreed to insert it verbatim into paragraph 34 of the Tunis Agenda, the ultimate product of WSIS.

In the work of the WGIG, a large fraction of the time and energy was dedicated to investigating the position of ICANN, which was financially supporting the group. Not only was the sui generis organization discussed as a governance structure that did not give an equal standing to governments, but also as an embodiment of a privatized way of operating that needed to (p.123) become more inclusive. As the Internet became widespread in other parts of the world, it could no longer have online content primarily in English and domain names could no longer be typed with Latin characters only. Support for multilingualism and local content was dominant. The final report released by the WGIG in July 2005 addressed the need for a restructuring of IG, but did not influence the status quo with regards to ICANN.

Drafted at Château de Bossey outside Geneva, with consideration for the opinions expressed during the open consultations, the report elaborated on an open forum function for all stakeholders and four proposals for public oversight arrangements. The members of the group did not reach agreement on whether one model would be preferable to the others, and Desai concluded that they all needed to be included as ‘four equally beautiful brides’ (Drake 2005). Three of these proposals envisioned an international body replacing ICANN or overseeing the Internet Assigned Numbers Authority (IANA) function (thus eliminating the control of the US Department of Commerce (DoC)), yet none was formally endorsed in the Tunis Agenda. One of the models discussed—an intergovernmental, UN-based ‘Global Internet Council’ (GIC)—replacing the DoC oversight and the ICANN Governmental Advisory Committee, garnered increasing attention in the following years, as discussed in the next sections of this book. In the various proposals put forward, the management of the root continued to be entrusted to ICANN or a globalized version of it (‘World ICANN’: linked to the United Nations with a host country agreement), yet the policymaking function would generally lie with a GIC, a Global Internet Governance Forum, or an International Internet Council. The non-state actors, both for-profit sector and civil society, were frequently side-lined in the models proposed, support being primarily shown for governments and for the United Nations (Raboy et al. 2010, 137).

In negotiating the oversight system, two main visions were consolidated (Kleinwaechter 2005). On the one hand, there were the supporters of the ‘do not fix what is not broken’ slogan—favouring the status quo and arguing for keeping ICANN unchanged. On the other hand, there were the proponents of a UN treaty, which would have allowed all governments to share responsibilities, oversight, and control. As the WGIG negotiations concluded in mid-June, the NTIA issued the US Principles on the Internet’s Domain Name and Addressing System (on 30 June 2005), making three important statements: first, that the US government intended to ‘maintain its historic role in authorizing changes or modifications to the authoritative root zone file’; second, that ICANN was the appropriate technical manager for the domain names system; and third, that IG should continue to be discussed in multiple fora, in which the United States ‘will continue to support market-based (p.124) approaches and private sector leadership in Internet development broadly’ (NTIA 2005).

Without the endorsement of the United States and allies, nor that of the business sector, the negotiations on an alternative to ICANN did not advance in the following negotiation rounds for WSIS-II. What became obvious through the work of WGIG was the understanding that the Internet was, above all, a political issue with many more stakeholders asking for a seat at the table. The way in which the governance of the Internet was articulated up until that point came under heavy contestation and that was reflected in the interventions at the WGIG open consultations. The latter became, in the process, ‘an “obligatory passage point” at a crucial stage in the ordering of the global politics of the digital revolution’ (Flyverbom 2011, 117).

Moreover, the WGIG introduced a set of procedures that remained emblematic for IG processes: it introduced a multi-stakeholder extra-budgetary funding scheme (further pursued in the constitution of the Internet Governance Forum (IGF))12 and it committed to a transparent process by holding open sessions and inviting open comments on draft documents (Mathiason 2009). One innovation that the WGIG adopted with inspiration from the ICANN meetings was the use of real-time captioning of the discussions in English. At the Château de Bossey, parts of the final report were drafted and fine-tuned collectively, using a projector so that all participants could work on the text at the same time (McLean 2005). The WGIG also validated the practice of setting up a working group to scope what was at stake, reconcile dissenting opinions and views, and propose ways forward. The practice of ad hoc groups in charge of finding a mediating position relied on legitimacy through expertise and, reiterated in multiple fora, became a defining routine in the IG field.

The WSIS Ordering

When the WSIS Summit came to an end in Tunisia in 2005 with the adoption of two outcome documents—the Tunis Commitment and the Tunis Agenda—the global ordering that was in flux during the negotiations authoritatively stabilized. The redefinition of the Internet as a sociopolitical field (in opposition to technical only) expanded the scope and exposure of what became a global public policy domain. A sustained public commons discourse surfaced in the contributions of the delegations from developing countries, (p.125) countered by an understanding of the Internet as a ‘global facility’ in the eyes of the business sector and endorsed by some of the participants representing Western governments. Focusing attention on the overarching theme of digital divide, WSIS-II discussed themes such as spam (fraudulent and unsolicited harmful email), multilingual access to the Internet, and the international Internet interconnection charges.13 These three points advanced throughout the PrepComs and as part of the WGIG consultations were arguably the most heated. A tripartite categorization of stakeholders (governments–businesses–civil society) hardened the positions of the key actors, each of which struggled with internal heterogeneity.

In the Tunis Agenda, governments agreed on the following distinction of roles and responsibilities of stakeholders, which continues to guide the current understanding of the shared competences in IG:

  1. a. Policy authority for Internet-related public policy issues is the sovereign right of States. They have rights and responsibilities for international Internet-related public policy issues.

  2. b. The private sector has had, and should continue to have, an important role in the development of the Internet, both in the technical and economic fields.

  3. c. Civil society has also played an important role on Internet matters, especially at community level, and should continue to play such a role.

  4. d. Intergovernmental organizations have had, and should continue to have, a facilitating role in the coordination of Internet-related public policy issues.

  5. e. International organizations have also had and should continue to have an important role in the development of Internet-related technical standards and relevant policies. (WSIS 2005a, paragraph 35)

The Tunis Agenda also recognized, in paragraph 36, the ‘valuable contribution by the academic and technical communities within those stakeholder groups mentioned in paragraph 35’. This further divided the core communities contributing to IG, ‘unilaterally relegating all other stakeholders to subordinate roles’ (Doria 2014, 124). Antagonizing these groups by relegating them to the side-lines, the Summit further called into question the extent to which technical standards and protocols were responsive to the needs of developing countries, in particular in promoting internationalized domain name and multilingual local content. The Tunis Agenda understanding—refining the wording of the Geneva Plan of Action from 2003—became a reference point in all subsequent discussions.

(p.126) In its IG section, the Tunis Agenda also reflected on the promotion, development, and implementation of a global culture of cybersecurity (in cooperation with all stakeholders and international expert bodies) in paragraph 39, reiterating what the United States proposed in related discussions in the General Assembly (Radu 2013). In the words of Kummer, the ‘Tunis Agenda was a “diplomatic compromise”, the beauty of which is that it is full of creative ambiguity that allows everybody to find something to satisfy their own wishes. As the Agenda was based on a decision-making Summit, the text on controversial topics such as IPR was carefully balanced in a way that avoided going into details that could be divisive and difficult to resolve’ (cited in Dutton et al. 2007, 5). Using the common-denominator approach meant that no IG reform proposed by the WGIG was pursued, with the exception of the endorsement for the creation of an IGF14 to address Internet-related questions in an open format, with non-binding recommendations.

The IGF was aimed at building consensus and shaping the development of Internet-related policies in a bottom-up manner, also fulfilling a socialization function by bringing stakeholders together for open discussions. Paragraph 72 of the Tunis Agenda established this forum for multi-stakeholder policy dialogue to discuss public policy issues, facilitate exchange among interested parties, identify emerging issues, and contribute to capacity building in developing countries. Organized by a small IGF Secretariat15 based in Geneva and reporting to Department of Economic and Social Affairs (UN DESA), the forum has been convened annually in different locations around the world since 2006 under a UN mandate that was originally valid for five years, renewed for the first time in 2010. At the WSIS decennial review in December 2015, the IGF was renewed for another ten years.

The concession made on the thorny issue of government participation in IG processes is summarized in paragraph 69 of the Tunis Agenda:

We further recognize the need for enhanced cooperation in the future, to enable governments, on an equal footing, to carry out their roles and responsibilities, in international public policy issues pertaining to the Internet, but not in the day-to-day technical and operational matters, that do not impact on international public policy issues. (WSIS 2005a)

(p.127) This wording led to a dialogue track perpetuated in what was known as the ‘Enhanced Cooperation process’. Different reactions were triggered by this formulation: for some stakeholders, an active participation in the IGF met the conditions for an enhanced cooperation. This is why most of the stakeholders reporting to the UN Secretary-General mentioned their efforts in the IGF process as their contribution to an enhanced cooperation in IG. For others, the IGF and enhanced cooperation represented two different processes. It is the latter view that the UN Economic and Social Council (ECOSOC) endorsed when the issue was brought to the fore in 2012.

Once again, behind the separation or unification of the two processes stood a firmer decision on the role of the UN system in IG. The enhanced cooperation dialogue was mandated by the Secretary-General to his special adviser, Desai, tasked to report on progress annually. Subsequently, the process was assigned to the UN Commission on Science and Technology for Development (CSTD), which started conducting discussions in the framework of the CSTD Working Group on Enhanced Cooperation (WGEC) chaired by Peter Major of Hungary, active starting in 2013. The group comprised twenty UN member states and five representatives from each of the four identified stakeholder groups (business; civil society; intergovernmental organizations; technical and academic community). Discussion on the ‘enhanced cooperation’ track soon stalled over the definition of terms.

While the role of governments at the global level was not settled at the WSIS, a tacit agreement was reached that countries retained their rights to legislate nationally (Hill 2014). The Tunis Agenda did not debate the mechanisms already in place at the national level, be they legislative or executive. In terms of new mechanisms, the WSIS consensus was limited to encouraging the establishment of a national implementation instrument that would contribute to the achievement of the MDGs. Mainstreaming links between ICTs and development at various levels, the Tunis Agenda called on regional intergovernmental organizations to work with other stakeholders to implement activities, exchange information and share best practices.

Role of the United Nations

The WSIS permanently transformed the IG field by crystallizing the UN participation in its processes. First and foremost, it clarified the facilitator role to be played by a number of UN agencies. In the Tunis Agenda, responsibilities for action lines were entrusted to different UN agencies: UN DESA—development, e-government, international cooperation; ITU—cybersecurity, infrastructure, enabling environment, capacity building; WIPO—copyright; (p.128) UNESCO—access, science, e-learning, cultural and linguistic diversity, media, information ethics. Through WSIS, various UN organizations defined their interests more clearly with regard to the Internet. ITU, WIPO, and UNESCO linked their mission directly to the Internet, whereas bodies such as UN Institute for Disarmament Research (UNIDIR) or UN Office on Drugs and Crime (UNODC) saw the Internet policymaking debates as tangential to their mission.

Second, a formalized reporting system was set in place. Assigning responsibilities created a debate, with concerns being put forward regarding the very definition of the implementation and follow-up mechanisms for WSIS. Whether the process would result in new institutions or expanded mandates for certain international organizations and governments, and whether multi-stakeholder participation would be preserved were highly disputed. The need for a reporting and implementation system was stressed in the recommendations of the Ad Hoc Working Group of the GA, tasked to examine the integrated and coordinated implementation of and follow-up to the outcomes of the major UN conferences and summits on socio-economic topics. The work of the ad hoc group happened around the same time as WSIS-I and resulted in the UNGA 57/270B resolution adopted in July 2003, recognizing the responsibility of the UN system to assist governments to stay fully engaged in the stocktaking process.

The WSIS stocktaking process started in August 2004 following a decision taken at the first meeting of Phase II preparations for Tunis. Given the overlapping mandates of several UN bodies, the stocktaking exercise was seen as an opportunity to coordinate and harmonize efforts across the UN system. To arrive at a full picture of the different activities and responsibilities undertaken by various entities, the ITU Secretary-General, in coordination with the WSIS Executive Secretariat, sent a letter to WSIS participants to invite them to fill in a questionnaire regarding the field of activities and actions they contributed to in line with the eleven themes of the Geneva Plan of Action.

The negotiation of follow-up measures was as political as the definition of IG, going from the proposal of concrete implementation activities and the creation of a new organizational mandate to coordinate these to the vague text of ‘a request to the Secretary-General of the UN to submit a report on implementation activities of the WSIS decisions’ in a draft from August 2005. After a new round of negotiations, an agreement was reached to entrust ECOSOC with overseeing the system-wide follow-up. ECOSOC’s subsidiary body, the CSTD was further mandated to evaluate progress on implementation and follow-up and to propose initiatives for improving their efficiency. In this process, the CSTD—with an agenda, mandate, and composition decided on (p.129) by ECOSOC—was named focal point and given the ultimate responsibility over the follow-up process.

The ITU, on the other hand, remained in charge of the stocktaking process, following the organization of the first meeting on this topic in October 2004. In the understanding of the time, stocktaking consisted of compiling the information regarding the different activities and initiatives undertaken by various actors in connection with the themes of the Summit. Participation in this exercise was voluntary, and the ITU took the lead in publishing an annual report based on data related to the themes identified in the Geneva Plan of Action. The Tunis Agenda also introduced a quantification approach with strong reliance on data, indicators, and indexes for monitoring the evolution of the information society. It pointed to the periodical evaluation of problems identified at the WSIS, noting that different levels of development and national circumstances need to be taken into account. The Partnership on Measuring ICTs for Development was one of its outcomes, providing a measurement of the efficiency of the implementation measures through the ICT Opportunity Index and the Digital Opportunity Index. Building on paragraphs 109 to 110 of the Tunis Agenda, the ITU organized, together with UNESCO, UNCTAD, and UNDP, the annual WSIS Forum, the largest global meeting for ICT4D, annually held in Geneva.

On the question of coordination and harmonization of activities in the UN system, the Group on the Information Society (UNGIS) was established by Kofi Annan in April 2006 within the UN Chief Executives Board for Coordination (CEB). The UNGIS consisted of twenty-eight UN bodies and organizations facilitating the implementation of the WSIS outcomes and coordinating the mechanisms for national and international implementation established in the Tunis Agenda. They initiated a process of annual meetings to identify priorities and coordinate activities and joint initiatives, consolidating cross-sectoral cooperation and strengthening the role and visibility of the UN on information society matters.

At the end of the WSIS, it became clear that the institutionalization of an emerging and highly politicized field was well underway, with the involvement of the regional and international organizations. Not only was the Internet fully integrated in global governance processes, it also became a visible point of contention between developing and developed countries over participation in the creation of rules for the Internet, which progressively blurred the boundaries between traditional sectors of regulation. The process laid the foundations for a developmental angle to be added to the evolution of the Internet. The organization of the WSIS brought together different UN bodies working at the intersection of technology and development, it consolidated the role of the ITU, and mandated follow-up procedures though which (p.130) a number of organizational mandates were extended. The Tunis outcome documents and UN General Assembly Resolution 60/252 further resolved to conduct an Overall Review of the Implementation of the WSIS Outcomes in 2015, a process discussed in the Post-Snowden Fault Lines section below.


Soon after the conclusion of WSIS-II, the ITU met for its 2006 Plenipotentiary meeting in Antalya and decided, among others, to convene a World Conference on International Telecommunications in 2012 (WCIT-12) to discuss the revision of the 1988 International Telecommunication Regulations (ITRs) treaty in light of the changes in the international environment. The ITRs served as a global treaty binding signatories to comply with general principles for worldwide interconnection and interoperability of existing communication services and facilitate the availability of international telecommunication services and networks. It said little about the Internet, which was in its infancy at the time.

In the lead-up to the WCIT-12, core frictions emerged between stakeholders and interest groups on state involvement in Internet regulation. The United States16 and many supporters of the multi-stakeholder practice voiced concerns that the revised ITRs posed threats to an open Internet and called for opposing the alleged UN intent to govern and regulate the Internet and related attempts by member states to impose governmental leadership via this treaty. On 22 September 2012, the US House of representatives unanimously approved a resolution urging the US government not to give the ITU control over the Internet. A broad coalition of non-state actors, including civil society organizations and large Internet companies, mobilized against the WCIT-12.

Amidst these tensions, the ITU Secretary General Hamadoun Touré (2012) repeatedly stated that the WCIT-12 would not address the Internet, yet a number of documents among the 1,275 revision proposals submitted by member states referred directly to the Internet. The Russian proposal was worrisome to many: ‘member states shall have equal rights to manage the Internet, including in regard to the allotment, assignment and reclamation of Internet numbering, naming, addressing and identification resources and to support for the operation and development of the basic Internet infrastructure’. The talks around amending the ITRs were also prominently (p.131) contested for increasing the jurisdiction and legal control of the ITU over the Internet (Blackman 2013). The WCIT-12 included, apart from the revision of the treaty, discussions around the adoption of a non-binding resolution for fostering an enabling environment for the greater growth of the Internet, which affirmed a strong and continuous role for the ITU in IG, stating that ‘all governments should have an equal role and responsibility for international Internet governance’.

The concern that dominated the meeting—the potential intergovernmental control of the Internet—was by no means new. Since 2011, non-Western countries put forward various proposals to bring network governance into state-led global or regional frameworks: India, Brazil, and South Africa pushed for the creation of a UN Committee for Internet-related Policies (CIRP), while Russia, China, Tajikistan, and Uzbekistan called for an International Code of Conduct for Information Security, drawing on an older Russian proposal submitted to the UNGA on information security (Radu 2013). The latter included a call for the creation of ‘a multilateral, transparent and democratic international management of the Internet’, reiterating the WSIS wording from negotiation drafts. The ITRs conference revived this discussion around intergovernmentalism versus multi-stakeholderism in the governance of the field and divided participants over the extent to which the treaty covered the Internet (Radu et al. 2014).

When more than 1,600 delegates from 151 member states met in Dubai on 3 to 14 December 2012 for the UN-sponsored meeting to amend the ITRs, the talks epitomized ideological struggles and conflicts between different socio-economic and representation models. The text of the treaty was satisfactory to many developing countries, as it covered, among others, means to foster transparency and competition for international mobile roaming charges, to reduce electronic waste, increase accessibility to international telecommunication services for persons with disabilities, and ways to facilitate access to international optical fibre networks for landlocked and small island developing states. At the same time, the preamble made reference to the commitment of member states to uphold human rights obligations in the implementation of the new ITRs.17 The stakeholders opposing the WCIT negotiations questioned not only the legitimacy of the ITU to oversee security-related issues, in particular in the context of the provisions regarding cybersecurity (Article 5A) and unsolicited bulk electronic communications (Article 5B), but also the fact that participation in the discussions was (p.132) relatively limited (observation and advisory role) and voting was restricted to states only.

In the revised ITRs, Article 5A bound states to ‘endeavour to ensure the security and robustness of international telecommunication networks’, which raised concerns about limitations on content that signatory countries might pursue, thus legitimizing potential infringements on human rights. In this case, the negotiations revealed older tensions between the approaches of Russia, China, and the United States on information security, which previously surfaced in the UN General Assembly. The revised Article 5B of the treaty, stating that ‘Member States should endeavour to take necessary measures to prevent the propagation of unsolicited bulk electronic communications and minimize its impact on international telecommunication services’, was seen as a possible avenue to interfere with content online, for example in order to limit freedom of expression. Similar concerns surrounded the wording of the preamble, which contained a provision upholding the right of access of member states to international telecommunication services.

Given these fault lines, consensus was impossible to reach at the WCIT-12 and majority voting was used instead. Out of the 144 delegations with voting rights, eighty-nine were in favour (including Russia, China, Arab states, Iran, Brazil, Argentina, Mexico, Indonesia, South Korea, Turkey, and many African states) and fifty-five were against (including the United States, the United Kingdom, Canada, EU member-states, Australia, Japan, India, Kenya). The possibility was left for other states to join the new treaty later. Figure 2 shows the block of countries opposing the treaty, revealing a split along the developed–developing country dimension in the vote of the national delegations.

The WSIS Decade and the Public–Private Partnership Thirst

Figure 2 Non-signatories of the 2012 ITRs

The split in the votes was described as a ‘digital Cold War’ (The Economist 2012; Mueller 2013), with countries divided into two camps led by the United States and Russia. A similar grouping of states could also be observed five months after the WCIT, on 16 to 18 May 2013, when the ITU hosted the World Telecommunication/ICTs Policy Forum in Geneva, a non-binding meeting facilitating exchange and information sharing on emerging telecommunication regulatory matters for states and sector members. The revised 2012 ITRs entered into force on 1 January 2015, allowing for the new rules and the 1988 ones to operate alongside. Against this background, a number of IG processes (inter alia, the decennial review of the World Summit on Information Society) were perceived as potential battlefields for the definition of future IG mechanisms.

In the ICANN space, 2012 is remembered for the new expansion of the DNS. Applications for the new generic top-level domains (gTLDs) were invited in January that year. Unlike in previous rounds, new domains could (p.133) (p.134) be put forward in scripts other than Latin as Internationalized Domain Names, and be used for a variety of purposes, from business to community-oriented, including brand names or geographic strings. Policies to guide the application process had been in the making since 2005 in ICANN’s Generic Names Supporting Organization, but the launch of the applications was met with a lot of criticism, this time by the private sector. In 2011, a coalition of seventy-nine corporations organized under the US Association of National Advertisers signed a petition against the new gTLD programme. Since 2014, their interests were channelled by the Coalition for Responsible Internet Domain Oversight, comprising seventy-nine companies and 102 associations.

Notwithstanding concerns regarding the protection of trademarks in domain name registrations and fears of confusing and predatory registrations, the new programme attracted a total of 1,930 applications at the end of the registration period and ICANN held a prioritization draw in December 2012 to move forward with independent third party expert processing. By October 2013, the first four new gTLDs were delegated,18 followed by 475 more before January 2015. A string delegated in February 2015, .sucks,19 brought the debate to a new level, opposing freedom of expression entitlements to suspicions of exploitation and coercion of brand holders.

Numerous underlying tensions in the IG space came to the surface between 2010 and 2012. States- or corporations-driven apprehensions over Internet ruling profiled more clearly the new interests at stake, whether it was agenda-setting power, policymaking, or direct influence over institutional processes. The Pandora’s box was opened again in 2013, when confidential documents leaked by Edward Snowden showed the pervasiveness of the surveillance apparatus of the United States and its world-dominant national security agenda. The Snowden revelations brought to the forefront the cooperation between the US federal government agencies and Internet companies, overriding user privacy in pursuit of national interests. Beyond the specifics of the US intelligence-gathering practices under the PRISM mass surveillance programme, the disclosures marked a turning point in pushing for a more transparent and more accountable governance structure for the cyberspace.

(p.135) Post-Snowden Fault Lines

A deep rift formed in IG in mid-2013. On 5 June, the British newspaper The Guardian started publishing a series of articles about the secret collection of data by the US National Security Agency (NSA) and subsequently disclosed the wide scope of the PRISM surveillance programme. The data came from classified government documents leaked by Edward Snowden, a computer security specialist and former NSA contractor, who left the United States to seek temporary asylum in Russia. Three international media outlets—the US-based Washington Post, the British Guardian, and the German Der Spiegel—analysed the documents released by Snowden and reported that thirty-five world leaders had their phone communications tapped (including Brazil’s President Dilma Rousseff and German Chancellor Angela Merkel) and that the NSA surveillance capabilities had also been used in monitoring the communications of foreign companies like Huawei in China and Petrobras in Brazil (Globo 2013). According to the Washington Post, the NSA collected around 5 billion phone records every day (Gellman and Soltani 2013a) and 97 billion pieces of data in March 2013 alone. Part of the information was collected by tapping directly into the unsecured connections of Google, Yahoo!, and Microsoft data centres around the world (Gellman and Soltani 2013b).

It was not the first time confidential documents from the US administration were made public, generating reactions worldwide. Prior to the Snowden revelations, Wikileaks held the headlines for most of 2010 by publishing footage from the 2007 Baghdad airstrike, complemented, a couple of months later, by the Iraqi War Logs. In July, they released around 77,000 documents compiled in the ‘Afghan War Diary’ and in November the organization leaked diplomatic cables of the US State Department, generating a massive outcry from political leaders. The documents were disclosed by Chelsea Manning (born Bradley Edward Manning), who was soon after convicted on multiple charges and served time at a maximum-security facility in Kansas, United States.20

Manning’s revelations showed the power of the Internet in disseminating information, yet Snowden’s classified documents revealed the global ramifications of the US surveillance apparatus. Alongside the NSA, the British Government Communications Headquarters and Communications Security Establishment Canada also engaged in extensive spying on individuals and (p.136) on political organizations. The US-driven discourse on Internet freedom, promoted in particular by the US Secretary of State Hillary Clinton lost its credibility when the Snowden leaks went to press. There was a sharp contrast between the free flow of information discourse and the practices of the NSA. Back in 2010, Secretary Clinton argued that:

Information freedom supports the peace and security that provides a foundation for global progress. Historically, asymmetrical access to information is one of the leading causes of interstate conflict. When we face serious disputes or dangerous incidents, it’s critical that people on both sides of the problem have access to the same set of facts and opinions. (Clinton 2010)

For the IG communities already divided over the roles attributed to each in the Tunis Agenda, the Snowden leaks provided an opportunity to call for reform, in particular to restore trust and increase collaboration towards transparency and accountability. The minimization of the role of the US government constituted one of the dividing lines, resulting in a transfer of the oversight role of the Department of Commerce over IANA. On 7 October 2013, the technical bodies in charge of managing global resources—ICANN, the IETF, the WWW Consortium, the Internet Architecture Board, ISOC, and all five of the regional Internet address registries—issued a common statement to call for ‘accelerating the globalization of ICANN and IANA functions, towards an environment in which all stakeholders, including all governments, participate on an equal footing’ (ISOC 2013). This was also echoed in an EU press release dated 12 February 2014, calling for the ‘establishment of a clear timeline for the globalization of ICANN and the IANA functions’, as well as ‘strengthening the multi-stakeholder model to preserve the Internet as a fast engine for innovation’.

Following the revelations, Brazil and Germany drafted a UN General Assembly resolution on ‘the right to privacy in the digital age’ adopted unanimously on 18 December 2013. The resolution affirmed that ‘the same rights that people have offline must also be protected online, including the right to privacy’ and called upon member states to review their procedures, practices, and legislation on the surveillance of communications, their interception and collection of personal data, including mass surveillance, with a ‘view to upholding the right to privacy by ensuring the full and effective implementation of all relevant obligations under international human rights law’ (UNGA 2013). The initiative also stirred a related process within the Human Rights Council, which appointed, in July 2015, the first Special Rapporteur on the Right to Privacy.

Building on its efforts in the UN, Brazil hosted the April 2014 Global Multistakeholder Meeting on Internet Governance, known as NetMundial, (p.137) announced after a tête-à-tête between the Brazilian President and ICANN’s President and CEO Fadi Chehadé. Prior to the meeting, a proposal was made to hold online consultations drawing on the successful experience of the Brazilian Civil Rights Framework for the Internet (Marco Civil), a law that was drafted collaboratively over many years using an online platform to gather inputs from a wide range of stakeholders. Upholding the multi-stakeholder principle, NetMundial was innovative in its approach to assigning an equal number of slots for the four groups of interests identified: governmental, business, technical and academic, and civil society, offering a rowing microphone for interventions in the plenary discussions.

NetMundial concluded with a negotiated final statement outlining IG principles and a roadmap for further action. The anti-surveillance focus was toned down in the negotiations (Radu et al. 2015), yet the document became emblematic as it recognized the evolving roles and responsibilities of stakeholders, a heated debate post-WSIS. Part of the discussions focused on ICANN and the NTIA announcement from the month before of its intent to finalize the privatization of the IANA functions, started in 1998. Following the meeting in São Paulo, the NetMundial Initiative was launched in Geneva in August that year by the Brazilian Internet Steering Committee (CGI.br), ICANN, and the WEF, hosted by the latter. The initiative pursued very broad objectives that did not translate in concrete actions and came under criticism for its limited legitimacy and opaque procedures around the selection of its steering committee participants. It subsequently refocused on capacity development and lost visibility.

Around the same time, preparations for the decennial review of the World Summit on Information Society (referred to as WSIS + 10) intensified. The fears that the process would be less open than it had been in 2003 and 2005 were confirmed as the negotiations remained primarily intergovernmental. With the involvement of a number of UN bodies, and prominently the background work of the CSTD, the decennial review was completed with an evaluation of the progress made since 2005 and a GA High-Level Meeting, held at the UN headquarters in New York in December 2015. The merging of the WSIS priorities and the Sustainable Development Goals (SDGs) agenda were in focus, presenting many similarities to the Geneva Plan of Action and Tunis Agenda. While ICT-related goals were not listed among the SDGs (with the exception of Goal 9c regarding improved access), the Internet was discussed as an enabling platform for achieving the Agenda 2030 goals.

Importantly, throughout the WSIS + 10 negotiations and in the submissions of the different national delegations, consensus emerged around four key issues: digital divide—mentioned by all but three of the submitters (ISOC 2015a)—Internet access, child online protection, and support for (p.138) multi-stakeholderism (broadly understood). While the first two issues represented areas in which significant progress was still needed, the latter were new items compared to the 2003 to 2005 WSIS discussions. The multiplicity of governance mechanisms acknowledged and evaluated ahead of the decennial review—stressing the multi-stakeholder component—revealed many instances of self-regulation and collaboration across sectors.

However, as in the past, the most heated discussion was around the ‘roles and responsibilities’ of actors, where no agreement could be reached. The final wording upheld the formulation included in the Tunis Agenda, recognizing that ‘the management of the Internet as a global facility includes multilateral, transparent, democratic and multi-stakeholder processes, with the full involvement of Governments, the private sector, civil society, international organisations technical and academic communities, and all other relevant stakeholders in their respective roles and responsibilities’ (paragraph 57). The long-negotiated outcome document, which received public input at different stages (with many contributions from civil society), made only one reference to surveillance in paragraph 46, as member states failed to agree on a more precise wording.

Mechanisms of Governance

The patchwork of international rules that emerged throughout the WSIS decade pointed to a clearer understanding of shared responsibility for Internet policymaking. More diverse and increasingly private sources of authoritative decisions were profiling, followed by regulatory responses at the national and international levels. Beside the hybrid configurations, a redefinition and scaling up of the public domain could be observed in the process. The much denser regulatory pressure during this period was driven by ideological, diplomatic, and technological developments. The governance arrangements thus formed—summarized in Table 5—relied much more on discursive and operative modelling than on hard law, making the voices of non-state actors better heard. Some of these were traditional players which acquired a significant responsibility for Internet regulation over the years; others were newcomers to the process and demanded self-regulation or an equal footing in the deliberations. As more international institutions introduced specialized mandates or expanded existing mandates to address the realm of IG, their political standing was called into question and status-quo maintenance drew supporters closer together. (p.139) (p.140)

Table 5 Governance arrangements (global and regional) for the period 2005–15 (based on a total of 212 instruments recorded in the dataset)





Legal enshrinement

Treaties, conventions


2009 SCO Agreement on Cooperation in the Field of International Information Security

2013 WIPO Marrakesh Treaty

2014 African Union Convention on Cyber Security and Personal Data Protection

Court judgments


2012 ECHR judgment Ahmet Yildirim v. Turkey

2014 CJEU judgment Costeja v. Google

2015 CJEU judgment Schems v. Facebook

Institutional solidification

Specialized bodies


2010 Broadband Commission for Digital Development

2005, 2010, 2013, 2015 UN Group of Governmental Experts on Developments in the Field of Information and Telecommunication in the Context of International Security

2014 Global Commission on Internet Governance

2015 Global Forum on Cyber Expertise

Strategic framework/ agenda


2007 Global Cybersecurity Agenda

2013 OSCE Confidence Building Measures (CBMs) Stemming from the Use of ICTs

2014 Facebook’s Free Basics

2015 Europe’s Digital Single Market

Monitoring and benchmarking


2011 Boston Consulting Group E-Intensity Index

2014 ITU’s Global Cybersecurity Index

2014 Internet Traffic Monitoring Data Sharing (TSUBAME project)




2009 CoE Prague Declarations on A New European Approach for Safer Internet for Children

2011 Deauville G8 Declaration

2013 UNESCO Sakhalin Declaration on Internet and Socio-Cultural Transformations

2014 UN Resolution on Right to Privacy in the Digital Age (A/RES/69/166)



2005 WB ICT in Education Toolkit for Policymakers, Planners & Practitioners

2008 GSMA Mobile Alliance Against Child Sexual Abuse Content

2011 Consumers International A Guide to Developing Consumer Protection Law

2013 CoE Model Framework on Net Neutrality

Globally, the WSIS decade can be seen as the ‘golden era of regulation’ (Levi-Faur and Jordana 2005) for the Internet. Governance instruments aimed at limiting and constraining—rather than merely shaping—the behaviour of other actors became the norm across all Internet subareas. The overall increase in the number of dedicated instruments for this period (212 recorded in the database, compared to seventy-four in the previous period) attested to the high salience of the field, which matured into an institutional domain open to political contestation. Modelling mechanisms employed during the WSIS decade—primarily discursive, but also operative—constituted 53 per cent of all formal instruments in use by relevant actors. In the wake of the 2013 revelations by Edward Snowden—evidencing the depth and extent of the NSA surveillance—the dominant position of the United States came under fire explicitly or implicitly in many statements. The operative measures that complemented the high-level speeches or declarations aimed at guiding the development of policies in developing countries or introducing soft law principles in the routines of various organizations.

In parallel, the rights agenda consolidated. Early references to it dated back to the 2003 WSIS Declaration of Principles which affirmed the importance of maintaining and strengthening rights—in particular in connection with development and freedom of expression. While that document remained vague about implementing this vision, later efforts to turn that into reality materialized first at the national level and subsequently at the regional and global levels. Access to broadband Internet was first declared a right in Finland in 2010. A year later, a report of the UN Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression, Frank La Rue, concluded that disconnecting people from the Internet was against international law and constituted a human rights violation (La Rue 2011).

(p.141) Time and again, fears of Internet fragmentation via data localization practices were also understood in terms of human rights: in response to the NTIA surveillance leaks, there were repeated calls by government leaders to keep their users’ data on national territory. The human rights framing, used extensively both before and after the Snowden revelations, grounded a number of discussions that cut across subfields, such as net neutrality. Ambiguously defined, the human rights discourse remained open to a plurality of political articulations, opening certain positions up for change and reinterpretation. This stood in sharp contrast with the securitization discourse solidifying in the cybersecurity arena around the same time.

With no mentioning of human rights, China’s prioritization of the ‘cyber sovereignty’ approach meant that everything happening on the networks inside its territory would be subject to national rule. By 2008, the decade-long development of the Great Firewall of China was completed and a new form of censorship was in place, limiting access to certain foreign websites and close control of cross-border Internet traffic. This was further enhanced via legislative action to stimulate the growth of the domestic Internet industry by limiting the presence of Western services21 or by the introduction of explicit provisions for foreign companies to respect domestic regulations. The last step in the process of developing an own Internet was the implementation of the National Public Security Work Informational Project (or the Golden Shield Project) for online content monitoring via surveillance and filtering. Two policies implemented by Xi’s government on anti-rumours and real name registration brought an ever-tighter control to the 560 million Chinese Internet users recorded in 2015. Harsh content controls were also introduced across the Middle East following the wave of Arab Spring protests in 2010 to 2012 and further expanded into authoritarian states in both Africa and Asia, in most cases implemented on Western technology (Deibert 2015).

Against this background, the strong emphasis on placing the Internet high up on global agendas and in strategic frameworks (14 per cent of all instruments recorded in the dataset) derived primarily from the creation of a global market based on the free flow of information. Improvements in connectivity speeds and mobile Internet made online transactions faster and more convenient. Be it for raising or disbursing resources, the Internet became a top global priority. Paradoxically though, financial mechanisms attached to Internet initiatives have varied in their success and the global Internet agenda has mostly been built discursively. The SDGs, part of the broad (p.142) intergovernmental UN framework known as Agenda 2030, did not include the Internet22 among its seventeen goals, but considered it a key enabling platform for achieving the targets.

At the end of 2005, half of the world’s population—mostly in developing countries—was not yet connected to the Internet (ITU 2015). Programmes like Facebook’s Internet.org (later renamed Free Basics), initiated in 2014 to connect hard-to-reach populations in developing countries at no cost for the user, saw a halt and a redesign amid protests for limiting access to only a few Facebook-sanctioned services online and violating net neutrality via its zero-rating agreement23 with mobile operators in the target areas. Many activists accused the company of introducing a ‘two-tiered Internet, on which new users could get stuck on a separate and unequal path to Internet connectivity, which will serve to widen—not narrow—the digital divide’ (Hern 2015); Facebook’s main competitors did not back the initiative. The programme was banned by the Indian regulator on net neutrality grounds in February 2016; however, it expanded and became operational in a number of African and Latin American countries (Zambia, Tanzania, Ghana, Kenya and South Africa, Colombia and Bolivia).

Like the Indian example shows, national regulation became overall more powerful vis-à-vis corporations towards the end of the WSIS decade, though wide differences persisted in terms of local regulatory capacity. The other route of Internet decision-making implementation was via the judiciary. In this period, 9 per cent of all authoritative decisions about the Internet came from courts. Gradually, the role of the judiciary in Internet policymaking strengthened, culminating with two key decisions of the Court of Justice of the European Union (CJEU) at the end of the WSIS decade: the introduction of the ‘right to be forgotten’ in 2014 and the 2015 invalidation of the Safe Harbour scheme for data transfers.

In the landmark case C-131/12 Google Spain SL, Google Inc. v. Agencia Española de Protección de Datos (AEPD) and Mario Costeja González, the CJEU established that, based on their right to erasure, individuals can ask for their personal information to be de-listed from search engines. Accordingly, it imposed an obligation on Google to act on the potential removal from search results of items that are ‘inadequate, irrelevant or no longer relevant, or excessive in relation to the purposes for which they were processed and in the light of the time that has elapsed’ (CJEU 2014), when the person concerned so requests.

(p.143) In practice, the CJEU judgment delegated to the company the case-by-case, human review of links for potential deletion, entrusting Google with a first decision on the matter; in case of dissatisfaction, the complaint could go to court, thus having the search engine’s decision as a first step in a longer legal process. In the implementation of the ‘right to be forgotten’, Google acquired quasi-judicial powers, adjudicating on the display of personal information, which constituted the core of its business model (Chenou and Radu 2017). To comply with the court decision, Google opened an online delinking request form on 29 May 2014 and received 12,000 deindexing requests in the following 24 hours. From September to November 2014, it held consultations with the newly established Advisory Council (made up of invited experts) in seven European capitals. One year after the CJEU judgment, Google received 253,258 de-listing requests with a total of 918,699 links to be evaluated. Of these, 58.7 per cent were not removed.

In 2015, EU’s highest court decided that the Safe Harbour agreement—which regulated data transfers to non-EU countries based on a guarantee of ‘adequate protection’—was invalid, since it failed to ensure it in the case brought forward by the young Austrian lawyer and activist Maximilian Schrems against Facebook Ireland Ltd. His complaint, filed with the Irish Data Protection Commission, followed the suspected involvement of the company in the PRISM mass surveillance programme and requested the prohibition of further data transfers from Ireland to the United States. The CJEU decided on 6 October that government interference and lack of legal remedies for individuals who seek to access data about themselves on specific services meant the United States had lower privacy standards than the EU, thus triggering negotiations around a new transatlantic data transfer scheme.

The value-laden rules imposed in the two Court cases discussed showed the power of courts in reversing authoritative decisions from earlier periods of time, be they company practices or transatlantic agreements. With an embryonic Internet industry, the EU redirected its efforts towards carving out new spaces for regulation in data control and data protection (Radu and Chenou 2015). Europe’s Digital Single Market, announced by the Juncker Commission in May 2015, aimed to create EU-wide regulation for e-commerce, telecommunications, and digital marketing, eliminating the national barriers and rendering the European market first in digital economy. With over 500 million Internet users with the highest purchasing power around the globe, the EU was, around the late 2000s, well-positioned to assume a strong digital regulator role by designing legislation and norms that were gradually adopted by many more countries outside its jurisdiction.

Governance throughout the WSIS decade was thus enacted though structural elements that directed, constrained, or defined action and partnership (p.144) frameworks. Although not in focus in previous decades, efforts to closely monitor and sanction came to characterize the period in which the Internet became a truly global medium. Such developments revealed that the enforcement capabilities generally associated with international institutions did not vanish, but were in fact complemented by benchmarking initiatives and comparative rankings.

Variation of governance mechanisms across subfields was a constant in the data analysed here. A clear focus on two areas of regulatory action emerged throughout the WSIS decade. Figure 3 provides an illustration of the thematic grouping and distribution of instruments in the six areas of interest analysed throughout this study. One-third of the instruments recorded deal with security, primarily cybersecurity, child online protection, and cybercrime. Another third address civil liberties, in particular privacy, data protection, and freedom of expression. Often presented as a trade-off, cybersecurity and online personal freedoms remained top of the agenda following the 9/11 attacks in the United States and subsequent attacks in Europe. The frequency and high impact of online disruptions took centre stage in the work of many institutions. The ITU launched its Global Cybersecurity Agenda in 2007, at the same time as its cyberpeace discourse, which lost visibility shortly after.

Internationally, the first cyberattack to paralyse a country happened in 2007 in Estonia, when governmental and bank services were inaccessible for several days due to distributed denial of service attacks, allegedly coordinated from Russia. In 2010, during the Russian–Georgian war over the independence status of South Ossetia, online attacks blocked the Georgian channels (p.145) of communication at strategic points in time (Radu 2013). Pertinent cyber actions throughout the WSIS decade could range from probing the limits of cyber defence at the state level or in the private sector, to signalling power positions, and finally to inflicting damage. At the global level, responses came primarily under the form of ad hoc security governance networks and public–private cooperation. Existing transnational governmental networks—including the North Atlantic Treaty Organization (NATO), the Shanghai Cooperation Organization, and the G8—started developing dedicated programmes and units; new forums for discussion have also emerged: the 2011 London Process initiated the biennial series of Global Cyberspace Conferences and led to the creation of Global Forum on Cyber Expertise (GFCE);24 the Munich Security Conference has included dedicated cyber activities since 2012.

Many other governmental initiatives focused on a broad understanding of cybersecurity and national capabilities, in particular as more and more Computer Emergency Response Teams (CERTs) and Computer Security Incident Response Teams (CSIRTs) were established for national-level coordination. Originally, the threats posed to Internet security were solved informally, without making appeals to other institutions, due to the localized nature of risks and their low impact. This changed with the rise of cross-border attacks and the limited expertise available, primarily concentrated within firms and closed communities of scientists (Schmidt 2014). While companies were often called in to work alongside state institutions in cyber crises, only governments could legally pursue perpetrators in the few cases in which attribution could be assigned. While intergovernmental cooperation on cybersecurity increased, other governmental initiatives, such as the Freedom Online Coalition, profiled themselves as fighting the downsides of securitizing and monitoring information flows. It is noteworthy that there was often overlapping membership in some of these divergent initiatives.

The civil liberties focus, captured in the human rights agenda and reinforced after the Snowden revelations, materialized in concrete efforts to design internet rights charters, such as those developed by the Association for Progressive Communication and the Internet Rights and Principles Coalition in 2013, which merged concerns for freedom of expression, access, and accessibility with privacy and security protections, as well as political participation and cultural and linguistic diversity. Soft regulation and self-regulatory tools also included significant references to rights, in particular at the level (p.146) of guiding principles. Going in the opposite direction, the CJEU decision against Google on the ‘right to be forgotten’ aimed at regulating dominant actors’ behaviour in the market (Chenou and Radu 2017) by imposing an obligation on them to de-list information when requested to do so. This ‘reversibility of authority’ that Hall and Biersteker discussed back in 2002 was introduced via courts, but also via soft law.

In this shift to tackle the Internet economy, the discursive repertoires of various actors became unstable over time, resulting in institutional ambiguity and longer strategic negotiations in the framework of binding instruments. As in the case of cybersecurity, the concentration of efforts on digital economy showed that the new political tensions were no longer about infrastructure, but about the market value of the Internet. Admittedly, in the struggle between states and the ‘Internet giants’, regulatory measures for information intermediaries have included a ‘turn to infrastructure’ (Musiani et al. 2015). Noteworthy cases include code remedies for interoperability.25The non-linearity of regulation, pushing certain priorities up and others down, made institutional ambiguity a permanent feature of the WSIS + 10 process. Global strategic agendas were reduced to the minimum common denominator to defer controversial choices.

The WSIS Decade and the Public–Private Partnership Thirst

Figure 3 Variation of governance mechanisms across subfields (2005–15)

Note: total does not add up due to multiple cases in which an instrument covers several subfields

The privileged role of the US administration continued to stir tensions. The legitimacy of the US government to set global rules via extraterritorial reach was questioned multiple times during the decade. In 2012, when the US Congress discussed two anti-piracy bills that triggered global protests, the Stop Online Piracy Act (SOPA) and the Protect IP Act (PIPA), the first global ‘Internet strike’ was organized on 18 January 2012. Over one billion Internet users worldwide (Sell 2013) were affected by the blackout of websites, consisting of temporarily making their content unavailable and redirecting users to a message opposing the proposed legislation. At stake were the regulatory over-reach powers granted by these laws and their wide-ranging consequences on Internet users worldwide.

In 2013, Microsoft challenged in court the request of the US government to turn over the emails stored in a data centre in Ireland of a target account linked to a drug investigation, opening a long-lasting inquiry into the extraterritoriality of law enforcement based on the 1986 Stored Communications Act. Meanwhile, Microsoft announced that it would open cloud servers in partnership with Deutsche Telekom in Germany, where the privacy laws are strict, to avoid warrants of a similar kind. The question of jurisdiction also (p.147) continued to surround the IANA stewardship transition process, announced in 2014; at stake was the California-based incorporation of the body created in 1998, in a previous NTIA effort to privatize the domain name system.

Moreover, the alignment of the US government with the Silicon Valley industry interests resulted in a push for regulatory action in the EU. In 2014, the EU Parliament discussed a non-binding resolution to break up Google, while the European Commission opened an antitrust investigation into Google’s search and advertising services. The role of private intermediaries was called into question as their financial and political power rose steeply. Their relation to governments was also probed, particularly from 2013 onwards. The digital business model based on the collection, storage, and third-party sharing of personal data, coupled with profiling for targeted advertising, was not the main target of regulation, but was subsequently affected by sectoral laws, such as those dealing with privacy and data protection.

As regulatory patterns evolved, the conceptualization of rule-making broadened. The use of modelling instruments became mainstreamed in the activities of organizations with tangential roles in IG. Importantly, during the WSIS decade and in particular post-Snowden, the authority of technical bodies expanded to public policies issues such as encryption and surveillance, whereas privacy and data protection became politically disputed. The continuous renegotiation of positions observed from 2012 to 2015 embodied a struggle over fundamental values and principles for global IG. In the absence of a central mechanism for coordinating Internet-related policies, hybrid configurations became the norm, building on the interconnectedness of actors.


The multiplication and decentralization of sources of authority during the WSIS decade brought about new challenges for legitimacy in IG, ranging from the protection of critical infrastructure to control over the domain name system. Not only did these challenges alter the position of bodies such as the ITU or ICANN, but they also defined fundamental choices for institutional trajectories. The WSIS decade activated a set of interests that solidified over the years and led to institutional mandate expansion, mission overlap, and alternative fora. Noteworthy, private intermediaries taking on authoritative stances, acting as defenders of human rights online, and calling on governments to protect individual freedoms were not uncommon in this space.

At the end of 2015, the United States was home to eleven out of the fifteen largest Internet businesses, ranked by market capitalization: Apple, (p.148) Google, Facebook, Amazon, eBay, Priceline, Salesforce, Yahoo!, Netflix, LinkedIn, and Twitter, followed by China, hosting Alibaba, Tencent, Baidu, and JD.com (Meeker 2015, 6). This dominance was closely mirrored at the IETF meetings, where the Americans and the Chinese became the best represented nationalities. This reconfiguration was closely linked to the growth of digital markets based on e-commerce and the development of new applications. The Chinese case is particularly interesting as it challenges the de-regulatory credo of innovation. Evolving in a tightly controlled space, the Chinese companies played a major role in reshaping the Chinese society, currently the largest Internet user base in the world.

Operating internationally, they also reshaped standardization procedures for the digital economy, along with other East Asian businesses like KT, Samsung, and LG in South Korea, or NTT, Panasonic, and Sony in Japan. The world’s largest e-commerce platform Alibaba (founded by Jack Ma), the search engine Baidu or Tencent, which developed the mobile messaging app WeChat, have not only enhanced competition in the digital market, but have also exposed new ways of operating under the restrictions of the Chinese government, which blocked services like Twitter, Facebook, WhatsApp, Viber, or Line. Moreover, Chinese companies have also restructured the manufacturing and infrastructure market, providing lower-cost options via Huawei or Xiaomi. Rather than undermining state authority, market players such as the ones discussed above often joined forces with governments.

More than a proof of polycentricity, the diversification of venues and governance mechanisms indicated a weaker consensus and stronger contention over key issues in IG, particularly after the Wikileaks and Snowden revelations. Concerns about the long dominance of US actors—governmental or private—in IG were articulated by both developing countries and non-state actors. In addition to the efforts by emerging powers to situate themselves more prominently in the field, alternative venues were created for discussing what did not fit the ‘mainstream’ discourse. Noteworthy examples include the alignment of positions within G77 and the creation of the Internet Ungovernance Forum (IUF) and the Internet Social Forum (ISF). The IUF has been held annually alongside the UN-driven Internet Governance Forum (IGF) since 2014, and focuses on Internet censorship, freedom of speech, surveillance, and privacy. It was first convened in Istanbul with the support of a few international (Association for Progressive Communication, Article 19, AccessNow, Open Rights Group, Tactical Tech, WebWeWant, etc.) and local civil society groups (Alternative Informatics Association, Istanbul HackerSpace, etc.). The ISF was a thematic grouping of the World (p.149) Social Forum (WSF)26 established in Tunis in 2015, proposing ‘Another (People’s) Internet’ and aiming to serve as a global space for ‘sharing information on endeavours and struggles for democracy, human rights and social justice in relation to the Internet, and developing collective action agendas’. Its founding document called for opposing surveillance, corporate dominance, and governmental abuse of power. Upon its creation, the ISF was endorsed by more than 100 civil society organizations.

While NGOs pushed for a more open environment and alternative opinion, some of the technical bodies remained confined to their traditional practices and were slow to respond to reforms. Within ICANN, in charge of the management of the Internet domain name and unique identifiers, the ‘virtually unconstrained power’ entrusted to its Board of Directors was first questioned at the beginning of the 2000s (Weber and Gunnarson 2012, 13), and subsequently in the work of CCWG-Accountability as part of the IANA Stewardship process. For the longest time, its decisions could not be reversed even in cases in which they contravened ICANN’s by-laws or other written commitments and they could only limitedly be contested. The new process of reform started with the transition of the IANA functions, which also included an accountability reform, making it mandatory for the organization to undergo changes before the oversight from the DoC was completed (originally expected to happen in September 2015, prolonged for another year). Other technical bodies started to integrate policy experts in their boards and expert committees, but their membership did not reflect this diversity: American and Chinese male technologists with private sector affiliations constituted the large majority.

Watershed moments such as the WSIS, the WCIT-12, the 2013 Snowden revelations, and NetMundial re-affirmed the strong position of governments and the divisions within this nominal grouping. There was a lack of agreement over security-related matters stemming from basic contestation over locating authoritative decision-making, thus calling into question the status-quo relationships among stakeholders. Russia and China refused to sign the CoE Cybercrime Convention, but proposed and signed their own information security agreement in the framework of the Shanghai Cooperation Group. The two powers, along with India, Brazil, and South Africa—also referred to as (p.150) the ‘BRICS’ grouping—have been very vocal about creating an institutional environment in which formal equality among states was observed and power differentials were offset. Their demand for neutralizing power imbalances in decision-making amounted to a procedural fairness argument, albeit one possible only in an intergovernmental arrangement.

Yet seeing this period solely as state-dominated is misleading. Rather, it is an exemplary embodiment of hybrid arrangements, which transform, in their application, the very subjects and objects of governance concerned. Not only have there been cross-sector partnerships, but also influential technological developments resulting from the cooperation of institutions generally perceived to be in competition, especially over the intergovernmental or multi-stakeholder approach. For example, UNESCO collaborated with ICANN on the internationalization of Internet domain names. The IETF worked with the ISO for developing standards, for example codecs to manage audio and video content online, but also with the ITU-T for the development of basic protocols (Chenou and Radu 2015). Moreover, the Tunis Agenda mandated a set of UN specialized bodies to jointly conduct the follow-up and monitoring of the process, either by initiating actions or by convening annual meetings for self-reporting purposes.

The blanket cyber-surveillance plan known as PRISM revealed by Snowden encouraged many civil society organizations and activist groups, whether based in the United States or elsewhere, to take action to raise awareness regarding privacy protections and encryption systems. Challenging norms of inclusion and representation (Hintz and Milan 2009), grassroots tech groups such as Anonymous started profiling themselves more visibly around web-defacing operations to contest Internet-related negotiations conducted behind closed doors. Public protests around free trade agreements led to the rejection of the Anti-Counterfeiting Trade Agreement (ACTA) in 2012 and to the halting of comprehensive trade agreements in their drafting phase (e.g. Transatlantic Trade and Investment Partnership or the Trans-Pacific Partnership). Such instances of contestation revealed the centrality of the Internet in broader economic governance, including the request to uphold the established IG principles in new forms of decision-making.

The power of various actors was interest- and mission-driven, but their functional positioning had to do with the capacity to take on roles not available to others in a densely populated policy space. The Tunis Agenda perpetuated the artificial distinction between public policy and technical issues in IG, de facto endorsing the private sector-led management of the Internet via its emphasis on maintaining the stability and security of the Internet in an unchanged configuration. Yet, leaving international organizations out of the tripartite categorization of stakeholders did not accurately reflect the (p.151) governance configurations emerging. By the end of the WSIS decade, we could distinguish among organizations with core, secondary, and tangential interests, be it in standard-setting processes or in regulating one or multiple areas pertaining to IG writ large. Technical bodies designed specifically for Internet standards and protocols or associated functions generally preserved that as their core mission.

Within these broad categories that place IG interests at the centre, there was variation over time as to where international organizations would position themselves. Bodies like the IGF Secretariat are designed with core IG interests in mind, whereas organizations traditionally involved in telecommunication regulation went through an expansion of mandates. It was the case for the ITU, but also for WIPO, which was among the first institutions to incorporate Internet-related matters in its daily work. Starting in the late 1990s, it was involved in negotiating different aspects of online intellectual property rights, as diverse as the 1998 Universal Domain Name Dispute Resolution Policy (UDRP) for domain names and the 2014 Marrakesh Treaty facilitating access to published works by persons who are blind, visually impaired, or otherwise print disabled.

For organizations with tangential interests, IG did not require a re-conceptualization of the work they were involved in. For example, UNESCO presented itself as playing a threefold role in IG: (1) contributing to legal, societal, and ethical aspects of policymaking; (2) participating in discussions and assisting in reaching long-term solutions in IG; (3) advocating for safeguarding key values like freedom of expression, cultural diversity, and openness in current and future Internet policies (UNESCO 2004). With a less sizable involvement, UNIDIR was the convenor of policy debates around cybersecurity and use of online resources for terrorist activities—carving out an IG space after the stalemate of disarmament negotiations.

The decade in which the Internet knew its largest expansion was set against the background of state engagement in the articulation of governance, both as part of trans-governmental networks and as part of international organizations. This period also federated the participation of non-state actors in international negotiations, with growing legitimacy bestowed on them, thus giving birth to large cross-sector coalitions. The hybrid nature of such governance configurations also permeated national delegations to international congresses, which—more often than not—comprised members of industry and civil society. Moves across sectors, in particular from government to business and vice versa, a phenomenon referred to as the ‘revolving door’, facilitated the spread of ideas in support of the status quo. High officials in charge of US information policies either started or completed their careers in the private sector. The US DoC and (p.152) the NTIA are telling examples in this sense, as they extensively consulted with the private sector in the development of policies on cybersecurity and innovation (McCarthy 2015).

Similar patterns of cross-sectoral moves could be seen with a number of early leaders involved in multiple IG fora, shifting from business gatherings to UN meetings. These patterns sometimes revealed an individual commitment to a particular issue; other times, engagement was circumstantial or position-dependent. In the civil society and technical and scientific community, the ‘multiple hat’ phenomenon was extremely common, as most influential individuals navigated between the public and the private sector in their careers and held several affiliations at a time (Chenou 2014).

Anchoring Practices: Ad Hoc Expert Groups

The increasing hybridization of governance processes during the WSIS decade played a pivotal role in laying the foundation for the habitual resort to ad hoc expert groups. They were supporting the institutionalization of initiatives specific to the Internet and signalling the key role of expertise in the construction of this issue domain. Formed on a case-by-case basis, ad hoc groups brought together, for a limited period of time, different stakeholders to develop and provide an expert opinion, generally summarized in a final report. The underlying tenet in the creation of ad hoc groups was the need to tackle problems as they emerged, in a piecemeal fashion, and legitimize courses of action and decisions in a technocratic manner. In some cases, establishing such a committee became a controversial process, but very few questioned the outcomes.

The practice emerged as a common solution in the work of the standardization bodies in the early days and its statute as a dominant practice post-WSIS in Internet-related policy issues at the end of the WSIS decade revealed the continuation of informal relations of authority. Out of functional necessity or in search for licensing one answer over many possible alternatives, ad hoc expert groups deployed expertise readily available. A key role continued to be played by individuals acting as authoritative community voices (Mathiason 2009). Influential members of the technical community (e.g. Vint Cerf) and academics-cum-practitioners who were active within and around the WGIG, such as Wolfgang Kleinwaechter, Bertrand de la Chapelle, Milton Mueller, William Drake, Jovan Kurbalija, or Jeanette Hoffman often received invitations to take part in high-level discussions and working groups reflecting on the future of the field (Chenou 2014). (p.153) Shaping processes such as the WGIG or the IGF, Nitin Desai and Markus Kummer continued to be at the forefront of UN-driven interactions with a plethora of stakeholders (Epstein 2012).

Despite the open nature in the operations of many of the groups formed at the time, concerns were raised from the outset regarding the authority entrusted in the newly formed committees and bodies. This was best illustrated by the heated discussions around the formation of the first Multistakeholder Advisory Group (MAG) for the IGF in 2006:

The MAG, an ad hoc creation of the UN Secretary General which is part of the WSIS decisions, was supposed to help the secretariat regarding IGF procedures, selection process to ensure pluralism and transparency, methodology—not predetermine content or agenda! This should be the task of the first IGF meeting itself—and the IGF should have final say on procedures as well. (Carlos Afonso cited in Raboy et al. 2010, 181)

As to the choice of establishing the MAG in the first place, Nitin Desai explained that the decision was influenced by discussions with donors. Answering a question from the floor during the February 2007 open consultations, Desai noted:

When you have a multistakeholder forum with everybody on an equal basis, the very process of constituting a bureau itself is problematic, but even more so when there’s no membership. It’s an open door. So then we clarified. We asked this question to the people that sponsored. And they said, ‘This is what we had in mind’. Because I said, ‘How do I constitute a bureau in an open forum?’ And then they explained that this is how it was supposed [to be resolved]. (Epstein 2012, footnote 106)

Complementing the emergence of focal points and venues for IG-related discussions, the creation of ad hoc expert groups became a dominant practice for overcoming impasse in IG processes. It followed in the footsteps of WGIG and was emulated across a number of institutions, from private to intergovernmental. In the work of the UN Global Alliance for ICT and Development (UN GAID), ad hoc groups took the form of Communities of Expertise. These were networks convened by UN GAID to bring together motivated and capable actors to address specific, well-defined ICT4D problems in a results-oriented manner and to identify and disseminate good practices. Initiated by a group of at least three organizations, such networks needed to be multi-stakeholder, but could take a number of organizational forms, from the more exploratory ‘green field’ initiatives to being based at the UN. Serving a variety of purposes, this dominant routine was generally limited to a functional mandate, narrowly defined.

(p.154) Within ICANN, the Country Code Names Supporting Organization (ccNSO) established an ad hoc Internet Governance Review group, whose mission was to provide an overview of the IG issues and discussions with relevance for the work of the cc-TLD community. The review group operated between 26 February 2014 and 12 October 2014. Similarly, the Council of Europe established an Ad Hoc Advisory Group on Cross-border Internet,27 tasked to draft a Committee of Ministers Declaration on Internet Governance Principles in 2011. The practice was also adopted by the ITU. Following the WCIT-12, at its 2013 World Telecommunication/ICTs Policy Forum in Geneva, an Informal Experts Group (IEG) was convened to approve the text of the draft opinions to the ITU Secretary General report. Unlike the other groupings discussed above, the IEG had an exceptionally large membership, with 117 participants from member states, thirteen from regional and other IOs, twenty-nine from civil society and industry, alongside other members from scientific and industrial organizations. In their work, six draft opinions to the report of the Secretary General were discussed, including support for multi-stakeholderism and enhanced cooperation.

Beyond the intergovernmental space in which prolonged negotiations and time-demanding procedures represented the norm, rather than the exception, this dominant routine also gained currency in the industry, to mobilize support around a cause. This was the case of the WCIT Ad Hoc Working Group, an industry-led coalition with members such as AT&T, Cisco, Comcast, Google, Intel, Microsoft, News Corporation, Oracle, Telefonica, Time Warner, Verisign, and Verizon, active in 2012 around the ITU Plenipotentiary meeting in Dubai. On that occasion, particularly compelling was also the example of the Ad Hoc Group on Internet-related resolutions formed by a group of delegates during the WCIT negotiations. Not-for-profit actors embraced the practice as a means to involve political figures and bring important social capital to their initiatives. In January 2014, the Centre for International Governance Innovation and the Royal Institute of International Affairs (Chatham House) launched a Global Commission on Internet Governance made up of twenty-nine members and chaired by Carl Bildt, Sweden’s Minister of Foreign Affairs, to work for two years on articulating a strategic vision of the future.

A characteristic of ad hoc expert groups is the autonomy in deciding how the work would be conducted. Some groups start with a narrowly defined (p.155) issue, while others call for a mapping exercise to understand what is covered and is not covered adequately by existing mechanisms or institutions. The modality of work usually includes face-to-face and online meetings. The ad hoc meetings are usually conducted under Chatham House rule28 to encourage free discussion and a more relaxed exchange of opinions. Moreover, depending on how sensitive the topic under discussion is, experts might prefer—for reputational concerns—to avoid a direct association of their names with what is publicly quoted.

Most of the expert gatherings established ad hoc would be convened by existing organizations or entities and have a political stake. Given their temporary nature, the working groups formed in a short time span would not challenge the position of the main actors in the IG field, nor would they propose radical reforms in terms of how the field was organized. The ideological struggle between a semi-privatized Internet model and an intergovernmental approach to it surfaced less and less in the reports issued by the ad hoc groups, despite the increasingly more sizeable mandates they were entrusted. An element of this was also reflected in the composition of the group, which was not open to public consultations. Senior members of the IG community and seasoned leaders with a broad understanding of the diplomatic and economic issues at stake constituted the large majority of ad hoc groups.

While the IG field had been at first dominated by computer scientists and engineers, subsequent developments around commercialization and litigation brought it closer to the community of entrepreneurs, investors, and lawyers. The structuration of the field was further enhanced with the contribution of the security community and with its opening up to a much wider audience following the WSIS. But crucial to the evolution of the field was the involvement of high-level figures from the political sphere, via explicit statements and positions, contributions to global agendas, and multifold participation in groups of experts, which became the established practice towards the end of the WSIS decade. Deployed as a meaning-making exercise, this anchoring practice led to iterative interactions shaping discourses and activities and supporting the enactment of constitutive rules by the most powerful. In this mutually enhancing exchange, the Internet became a public domain of action, whose global, regional, and national dimensions intersected.

(p.156) Synopsis

As a policy field coming of age, the Internet faced a more reflexive turn post-WSIS. Concerns for authority, legitimacy, and accountability—expressed by different stakeholders—become central to the consolidation of this governance domain. A number of challenges, stemming from three diverse sources, were embedded therein: first, questions were spawn by the modus operandi of the sui generis institutions of the field, such as the international technical bodies exercising public governance functions for the Internet, in particular in connection to ICANN; second, demands for a broader space for regulatory intervention resulted from the gradual adaptation of intergovernmental organizations and their closer involvement in global Internet policymaking via the WSIS process; third, and perhaps most importantly, the role of private intermediaries started to be questioned as their financial and political power exceeded that of developing countries. Their relation to certain governments was also probed, in particular after the 2013 surveillance revelations. Never before was the centrality of US-based institutions more contested than throughout the WSIS decade.

The expansion of the concept of IG also brought in actors that did not participate in the early days. Further classification, segmentation, and clustering of stakeholders ensued. The institutionalization of representation procedures and the clear categorization of stakeholder groups solidified and the contestation that surfaced rarely addressed the WSIS definition of IG. But the strict categorization of stakeholder groups into government, civil society, business, technical, and academic community did not match the reality of fluid and evolving, issue-specific groupings. Nor did it reflect the distribution of power among stakeholders.

Like many other institutional innovations, non-binding forums such as the IGF offered an experimental venue for testing how IG might be conducted with the participation of dominant and marginal actors, with core or tangential interests. It also offered a vessel for easing the tensions between the two dominant governance approaches at the time: intergovernmental and private. The involvement of civil society and industry actors at various levels, alongside governments, and the opening-up of new negotiation arenas was unprecedented during this decade. Emblematic of multi-institutional IG bricolage, these dimensions uncovered new patterns and agents of structural change, the addressees of such changes and the conflicting sources of authority. But what will become permanent and what will dissolve in the near future?


(1) MDG 8 called upon governments, ‘in cooperation with the private sector, [to] make available benefits of new technologies, especially information and communications’.

(2) The WSIS process also received input from the United Nations Information and Communication Technologies Task Force (UNICTTF) and the G8-created Digital Opportunities Task Force (DOT Force), where business interests were strongly represented.

(3) Attending the first PrepCom meeting were: 607 participants from 139 member states of the United Nations and of any specialized agencies; and observers. The latter category included: fifty-four participants from UN secretariat and organs, twenty-one participants from ten UN specialized agencies, thirty-five participants representing thirteen invited intergovernmental organizations, 223 participants from NGOs and civil society organizations, and thirty-four participants from business sector entities.

(4) The EU statement: ‘new mechanisms for governance at global and national levels encompassing a) issues related to the sector like electronic communications regulatory frameworks, data protection, network security and cyber-security, legal aspects of e-commerce and Internet governance as well as b) more general issues related to the new citizenship in the information age’.

(5) The Brazilian statement: ‘democratic and representative Governments should not be replaced by arbitrary groupings of private business and non-governmental institutions in decisions regarding the economic space brewing within powerful digital networks, such as the Internet. Organizing this new environment to the satisfaction of all, and ensuring the beneficial participation of developing countries and their societies is central to our work’.

(6) For increased network reliability, the anycast system was introduced in the early 2000s; there are currently more than 100 anycast instances of root servers all over the world.

(7) As McLean (2005) notes, the group was formally announced on 11 November, but decisions about the membership were made in advance and the participants were notified informally by the WGIG Secretariat ahead of time.

(8) It has been argued, however, that the process leading to the formulation of the definition was ‘not a multi-stakeholder effort’ due to the overrepresentation of states (some with repressive Internet policies) in WGIG (Raymond and deNardis 2015, 16).

(9) This was announced at the UNICTTF Global Forum on Internet Governance in New York, 24–25 March 2004.

(10) In the words of one of the WGIG members, Lyndall Shope-Mafole (South Africa): ‘The focus and the main reason for the establishment of this group and for asking the Secretary-General of the UN to establish this group is because the issue is not technical. I think the issue is political’ (statement made on 18 April 2005).

(11) This was arrived at through the ‘Inventory of Public Policy Issues and Priorities’ put forward at the first open consultation. It was a list of forty-six issues prepared with inspiration from a similar list made by the then vice-chair of the UNICTTF (McLean 2005).

(12) The main donors supporting the WGIG—the Swiss, Norwegian, and Dutch governments, together with the Swiss Education and Research Network and ICANN—became significant contributors to the IGF.

(13) To address this discontent, the Geneva Plan of Action encouraged ‘the creation and development of regional ICT backbones and Internet exchange points’ (2003, section C2).

(14) For Mueller (2010, 78), the creation of the IGF ‘was widely understood to be the kind of agreement that could get the WSIS out of its impasse; it allowed the critics to continue raising their issues in an official forum, but as a nonbinding discussion arena, could not do much harm to those interested in preserving the status quo’.

(15) It was originally funded through extra-budgetary contributions to a trust fund, with donations from governments, international organizations, and various stakeholders.

(16) At the meeting in Dubai, the US delegation was one of the most numerous, consisting of 100 individuals, of which forty belonged to the private sector and ten to civil society organizations.

(17) Paragraph 2 of the preamble reads: ‘Member States affirm their commitment to implement these Regulations in a manner that respects and upholds their human rights obligations.’

(18) The strings delegated on 23 October 2013 were: 游戏‎(xn--unup4y)—Chinese for ‘game(s)’, сайт‎ (xn--80aswg)—Russian for ‘site’, онлайн‎ (xn--80asehdb)—Russian for ‘online’, شبكة‎ (xn--ngbc5azd)—Arabic for ‘web/network’.

(19) The domain .sucks belonged to Vox Populi, a Cayman Islands based registry which allegedly charged a premium for brands to defensively reserve anything related to their name.

(20) Manning was released in May 2017, after her sentence was commuted by President Obama in January 2017 to 7 years of confinement since her arrest.

(21) China developed its own version of popular services: Youku (YouTube), Weibo (Twitter), Renren (Facebook), WeChat (WhatsApp).

(22) With the exception of Goal 9c, which refers to expanding Internet access.

(23) The zero-rating model consists of providing Internet access to certain websites free of charge as part of a user’s data plan.

(24) The GFCE acts as a global platform to share experiences, identify gaps, and strengthen cyber capacities worldwide. Its membership comprises fifty countries, international organizations, and companies.

(25) This is what the European Commission mandated for Microsoft in the final settlement of its antitrust case in 2009 and what the Federal Trade Commission decided in the antitrust settlement of Intel in 2010.

(26) The WSF was formed in opposition to the annual World Economic Forum (WEF) meeting in Davos, and its Charter calls for a different kind of globalization than that ‘commanded by the large multinational corporations and by the governments and international institutions at the service of those corporations’ interests’ (WSF Charter). The first edition of WSF was held in Pôrto Alegre in 2001.

(27) Its members were: Bertrand de la Chapelle, Wolfgang Kleinwaechter, Christian Singer, Rolf H. Weber, and Michael V. Yakushev.

(28) When a meeting, or part thereof, is held under the Chatham House rule, participants are free to use the information received, but neither the identity nor the affiliation of the speaker(s), nor that of any other participant, may be revealed.