Liberate printed matter, on the Web

View the Project on GitHub stevenjmesser/hypertext-store

Copyright © 2018 by the Centre for International Governance Innovation

The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the Centre for International Governance Innovation or its Board of Directors.

This work is licensed under a Creative Commons Attribution — Non-commercial — No Derivatives License. To view this license, visit (www.creativecommons.org/licenses/by-nc-nd/3.0/). For re-use or distribution, please include this copyright notice.

Printed in Canada on paper containing 100% post-consumer fibre and certified by the Forest Stewardship Council® and the Sustainable Forestry Initiative.

Centre for International Governance Innovation and CIGI are registered trademarks.

67 Erb Street West Waterloo, ON, Canada N2L 6C http://www.cigionline.org

CIGI Masthead


President Rohinton P. Medhora Deputy Director, International Intellectual Property Law and Innovation Bassem Awad Chief Financial Officer and Director of Operations Shelley Boettger Director of the Global Economy Program Robert Fay Director of the International Law Research Program Oonagh Fitzgerald Director of the Global Security & Politics Program Fen Osler Hampson Director of Human Resources Laura Kacur Deputy Director, International Environmental Law Silvia Maciunas Deputy Director, International Economic Law Hugo Perezcano Díaz Director, Evaluation and Partnerships Erica Shaw Managing Director and General Counsel Aaron Shull Director of Communications and Digital Media Spencer Tripp


Publisher Carol Bonnett Senior Publications Editor Jennifer Goyder Senior Publications Editor Nicole Langlois Publications Editor Susan Bubak Publications Editor Patricia Holmes Publications Editor Lynn Schellenberg Graphic Designer Melodie Wakefield

For publications enquiries, please contact publications@cigionline.org.


For media enquiries, please contact communications@cigionline.org. @cigionline

Table of Contents

vi About the Authors

vii About the Global Security & Politics Program

vii Acronyms and Abbreviations

1 Executive Summary

1 Introduction

2 The Creation and Governance of the Internet

3 Openness

5 The Geopolitics of Internet Ideals

12 Discussion: Four Internets and a Free Rider

13 Conclusion

14 Works Cited

17 About CIGI

17 À propos du CIGI

vi CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

About the Authors

Kieron O’Hara is an associate professor in
electronics and computer science at the University
of Southampton, UK. His interests are in the
philosophy and politics of digital modernity,
particularly the World Wide Web; key themes
are trust, privacy and ethics. He is the author
of several books on technology and politics, the
latest of which, The Theory and Practice of Social
Machines (Springer, with Nigel Shadbolt, David
De Roure and Wendy Hall), will appear in 2019.
He has also written extensively on political
philosophy and British politics. He is one of the
leads on the UK Anonymisation Network, which
disseminates best practices in data anonymization.
Dame Wendy Hall, DBE, FRS, FREng, is Regius
Professor of Computer Science at the University
of Southampton, UK, and an executive director
of the Web Science Institute at Southampton.
Her influence as one of the first to undertake
serious research in multimedia and hypermedia
has been significant in many areas, including
digital libraries, the development of the Semantic
Web and the emerging discipline of Web Science.
She became a Dame Commander of the British
Empire in 2009 and is a fellow of the Royal Society.
She has been president of the Association for
Computing Machinery, senior vice president of
the Royal Academy of Engineering, and a member
of the UK Prime Minister’s Council for Science
and Technology. She was a founding member
of the European Research Council and chair of
the European Commission’s Information Society
Technologies Advisory Group; a member of the
Global Commission on Internet Governance; and
a member of the World Economic Forum’s Global
Futures Council on the Digital Economy. Dame
Wendy was co-chair of the UK government’s
review of artificial intelligence (AI), Growing
the Artificial Intelligence Industry in the UK
(2017), and became the UK government’s first
“Skills Champion for AI in the UK” in 2018.
Four Internets: The Geopolitics of Digital Governance vii

About the Global Security

& Politics Program

The Global Security & Politics Program at CIGI focuses on a range of issues in global security, conflict management and international governance — a landscape that continues to change dramatically. Such changes are widely evident in the growing rivalry between China and the United States in the Asia-Pacific and the emergence of new economic powers in the region, such as Indonesia; the divergent ways Canada, Russia and the United States perceive Arctic security as melting ice opens up the Northwest Passage; continuing debates about the humanitarian imperative as the world confronts new crises in Africa and the Middle East; and new areas of concern such as cyber warfare and the security of the internet.

With experts from academia, national agencies, international institutions and the private sector, the Global Security & Politics Program supports research in the following areas: Arctic governance; Asia and the Pacific; fixing climate governance; governance of conflict management, with a focus on Africa; global politics and foreign policy; and internet governance.

Acronyms and


AI artificial intelligence
CNIL Commission Nationale de
l’Informatique et des Libertés
DNS domain name system
FCC Federal Communications Commission
GDPR General Data Protection Regulation
IETF Internet Engineering Task Force
IP Internet Protocol
IPv4 IP version 4
IPv6 IP version 6
ISPs internet service providers
SNSs social networking sites
VoIP Voice over Internet Protocol
W3C World Wide Web Consortium
Four Internets: The Geopolitics of Digital Governance 1

Executive Summary

The internet is not a monolithic architecture whose existence and form are guaranteed in perpetuity, but a fragile and contingent construction of hardware, software, standards and databases, governed by a wide range of private and public actors whose behaviour is constrained only by voluntary protocols. It is therefore subject to evolution and political pressure. Its original creators engineered it to be open, that is, that its standards should be transparent, and that data and software should be portable, extensible and interoperable. This Silicon Valley view was partly ideological, but partly based on engineering principles to enable the internet to scale as it grew. However, as the internet, and applications such as the Web, have become entrenched in daily life, competing views about how it should be governed have begun to emerge, and to be championed at the national level, where they are playing a geopolitical role. European nations, and the European Commission, envisage a “bourgeois” internet, where trolling and bad behaviour are minimized and privacy protected, possibly at the cost of innovation. Many nations, perhaps most notably China, see an authoritarian internet, where technologies of surveillance and identification help ensure social cohesion and security by combatting crime, terrorism, extremism and deviance. A more commercial view, characteristic of the US Republicans in Washington, DC, understands online resources as private property, whose owners can monetize them, exclude others from using them and seek market rates for their use. Finally, the openness of the internet is a vulnerability that can be exploited for misinformation or hacking, an opportunity taken by Russia, Iran and North Korea, among others. Thus, several internets are currently co-existing uneasily. We have not, however, reached an equilibrium; we need to be prepared for the internet that we know to evolve unpredictably, and work to ensure that it remains beneficial for humankind.


The internet is not a monolithic technological
creation, but a congeries of systems, protocols,
standards, hardware (the infamous “tubes”;
Blum 2012) and organizations. It encompasses
the domain name system (DNS), information
intermediaries, security systems, exchange points,
autonomous systems, internet service providers
(ISPs), registers, databases and standards bodies
— some with national standing, some (often in
the United States) with global reach, and others of
international standing — as well as some public
bodies, some private companies and some non-
profit organizations.^1 The system is truly socio-
technical — we cannot hive off the technical
from the rest. Every design decision reflects, and
imposes (perhaps unconsciously), a balance of
power, while cultural, economic and political
tensions play out across the collective-action
problems generated by digital modernity (O’Hara
2018). Neither computer science nor the social
sciences are individually sufficient to understand
this immensely complex piece of technology, the
structure of which is driven by the people who
upload content, download content and create the
links; the authors of this paper have long argued
that a dedicated “Web Science” is required both to
understand it and to engineer it (Berners-Lee et al.
2006; O’Hara et al. 2013; O’Hara and Hall 2013).
To complicate the politics, the internet grew out
of several US initiatives, and the United States
retains a disproportionate influence. However,
this position, which has fostered the growth
of the internet for decades, is under pressure.
International bodies have called for responsibility
for the internet to be transferred to more
international arenas — for example, the Working
Group on Internet Governance, under the auspices
of the International Telecommunication Union,
recommended in a 2005 report (paras. 52, 55)
that the United States relinquish oversight of the
system, the role ideally to be performed by a UN
body. The aim of such measures is to replace the
current ad hoc, decentralized, distributed model
of internet governance with a system of greater
legitimacy; the danger, however, is that such a
system would become centralized and sclerotic,
1 As described by Laura DeNardis (2014a), to whom this paper’s authors
are indebted for many insights.

2 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

focused on government power rather than on
the inclusion of, say, civil society or industry
voices. Arguments between democratic and
non-democratic states, for example, are likely to
dominate such a forum, with the risk of reducing
it to stalemate. Accordingly, such proposals have
struggled to find support, in part because few
doubt that the United States has, on the whole,
been a benign force on the internet and nurtured
its growth as few other nations could or would.
As a result, many commentators, by no means all
American, believe that the United States’ hands
would be far safer than the United Nations’.
More important than this diplomatic pressure
to change the system, therefore, has been the
application of power by various national and
supranational institutions to the delicately
balanced system itself, to try to “push” the internet
into a different type of model. This realpolitik is
having an effect, and it is clear that the internet,
as originally conceived by the primarily American
white male technologists who founded it, is
morphing into something else. But what?
The internet has many possible futures: it could
break or collapse under these pressures, as was
recently argued by Eric Schmidt, former chairman
of Google and Alphabet (Kolodny 2018); it could
develop unequally, with few if any benefits for
the half of humanity that is not connected; or it
could flourish over a diverse set of technologies
and geographical areas (Global Commission on
Internet Governance 2016). Progress at the moment
is equivocal. In this paper, the authors will argue
that four internets — at least — are emerging.
They are, at present, coexisting, and may continue
in this way for some time. It is possible, however,
that any of these internets may fall by the wayside,
and also that any one of them might become
dominant — or, indeed, that the whole intricate
system may collapse from these pressures.

The Creation and

Governance of the


In the context of clashing geopolitics, the internet
is a gossamer arrangement. Its core is the naming
system that gives one’s device a presence, an
identity, even a technical persona, on the internet
itself. Identifiers have to be globally unique and
universally accepted for the internet to function as
a global space. The main identifier is the numerical
Internet Protocol (IP) address (32 bits in length
in IP version 4 [IPv4], 128 bits in IPv6), which is
convertible by one of a number of recognized
organizations into the familiar domain name. The
link between IP addresses and domain names is
maintained by a hierarchically federated database.
The DNS is extraordinarily complex, with several
tasks to be coordinated in real time and at scale in
order that the essential system of unique naming
be preserved. Domain names need to be assigned,
and to be resolvable into IP addresses via the
database; the database needs to be edited and
maintained; the hierarchical naming structure
needs to be edited and maintained (for example,
authorizing new top-level domains on a par with
.com or .org); the servers containing this material
need to be operated and housed; new language
scripts beyond the Latin alphabet need to be
authorized and integrated; disputes need to be
resolved in a timely and legitimate fashion; and, not
least, the system needs to be secured from attack.
There are many other aspects to internet
governance, all equally complex, and requiring
intricate engineering and institutional coordination
across governmental and private bodies, and across
borders. The internet has a particular history, rooted
in Silicon Valley, and has been extraordinarily
successful. However, this history is contingent; it
could have been designed differently and it could
change over time. This paper will consider some
of the forces for change that are already altering
a structure that is sometimes taken for granted.
Four Internets: The Geopolitics of Digital Governance 3


Internet governance bodies are reflexively open. The Internet Engineering Task Force (IETF), which develops open internet standards, is highly participatory and transparent. Participation is not restricted by credentials, and the IETF’s documentation and records are open and freely available, allowing oversight and accountability. The IETF prefers to approve standards that do not rest on intellectual property and patents; where these do exist, it prefers royalty-free licensing. The World Wide Web Consortium, or W3C, has a similar policy with open standards and opposition to royalties, although it has a membership model and accepts institutions of any kind as members.

Openness of governance begets openness of technology. Until relatively recently, the operation of the network ignored the content of the packets of information that were routed around it. Routing algorithms applied to all packets indiscriminately, and the routers had no access to the content to derive grounds for discrimination. The headers of the packets, which contain the metadata about, for example, where the packets are headed, were the only things read. In this sense, there was no interference with the flow of information around the internet.

Support for this open system is often very idealistic. Yochai Benkler argues in his book The Wealth of Networks (2006, 131) that “the emergence of less capital-dependent forms of productive social organization [offers] the possibility that the emergence of the networked information economy will open up opportunities for improvement in economic justice, on scales both global and local.”

Hence, admiration of the technical brilliance of the internet design combines with an idealistic view of its affordances (what it, as an environment, offers the individual), a view which itself bifurcates. The admirer of the technology approves not only of the speed and efficiency with which data can travel from A to B, but also more generally of free speech, free association and other aspects of individual liberty. The result is a libertarian vision of the internet focused on its affordances, somewhat divorced from any messiness resulting from its collision with quotidian offline existence. On this view, the brilliant and elegant design complements the excitement of the freedom it

offers, and each becomes normative — the internet
should be free, because its design frees people
to develop authentically and autonomously.
The most famous statement of this philosophy
is John Perry Barlow’s “Declaration of the
Independence of Cyberspace,” which rejects
any idea that cyberspace needs real-world
institutions and remedies, arguing:
Cyberspace consists of transactions,
relationships, and thought itself, arrayed
like a standing wave in the web of
our communications. Ours is a world
that is both everywhere and nowhere,
but it is not where bodies live.
We are creating a world that all may
enter without privilege or prejudice
accorded by race, economic power,
military force, or station of birth.
We are creating a world where anyone,
anywhere may express his or her beliefs,
no matter how singular, without fear of
being coerced into silence or conformity.
Your legal concepts of property, expression,
identity, movement, and context do
not apply to us. They are all based on
matter, and there is no matter here.
...The only law that all our constituent
cultures would generally recognize
is the Golden Rule. (Barlow 1996)

Concerns about Openness

Until recently, it was assumed that this philosophy
of openness and liberty would carry all before it,
but many of the challenges facing authoritarian
opponents of openness 10 years ago (O’Hara 2009)
have been overcome. Benkler’s warm approval is
not the only possible reflection on the design of the
internet, especially when we think of that design
as a socio-technical construct rather than as a set
of elegant technical protocols. Most obviously,
the very idea of openness — in trade, migration,
capital movements and so on — is under threat
across the globe following the 2008 financial
crisis, and of course the open internet has been a
key part of globalization. Furthermore, openness
does not always guarantee equitable outcomes —
Silicon Valley has been called “a monoculture of
white male nerds” in which companies founded

4 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

by women received two percent of venture
capital funding in 2017 ( The Economist 2018s).^
In particular, central to the internet’s function
is that the key resources, notably the devices by
which the internet is accessed, are universal and
unique. A website needs the contacting device’s
IP address in order to deliver the requested
information to it (an IP address is regarded as
identifying in at least some circumstances by
the Court of Justice of the European Union^2 ).
Other unique identifiers are also vital for the
internet’s function, including various hardware
identifiers, cookies, real-name requirements
in social media, location information from IP
addresses and mobile base stations, and the
identification requirements associated with ISPs.
The uniqueness upon which the system depends
tends to spark three different responses.
The first is a worry about privacy. The function of the
democratic world depends on reasonable privacy
for individuals to consume news, political speech
or other cultural artefacts; to associate without
surveillance; and to organize action. The structure
of the internet holds out the possibility that people,
or at least their devices, can be traced, and their
downloads and uploads noted and recorded.
Attempts to solve this problem by using open
standards to allow users to express their privacy
preferences, such as the Platform for Privacy
Preferences^3 and “Do Not Track” mechanisms,^4
have not caught on nearly as well as blunter
instruments such as encryption standards, behind
which both the innocent and the guilty can shelter.
The second response, the converse to the first,
is to welcome an opportunity. Security and law
enforcement loom large in the responsibilities
of government, and the use of the internet as a
communication medium gives the prospect of
understanding criminal and terrorist networks
that threaten public security. Furthermore, the
patterns of interaction might also help to optimize
certain social functions, allowing government
intervention to use the data generated to improve
matters. The increasing prevalence of mobile
2 Patrick Breyer v Bundesrepublik Deutschland, [2016] EUECJ C-582/14.
3 Described at http://www.w3.org/P3P/ and now obsolete. It demanded rather a
lot of investment from users for only equivocal gains.
4 See http://www.w3.org/TR/tracking-dnt/; at the time of writing, this is a W3C
recommendation, but it remains unclear what exactly it means when a
user asks an application not to track her.
devices and the Internet of Things will extend the
reach of such benign (or not so benign) governance,
facilitating interventions in, say, health care and
well-being, climate change and traffic congestion.
A third response to the internet’s uniqueness is
not perhaps quite as obvious as privacy concerns
or the embrace of opportunity, but follows from
reflection about the business models that the
internet has ushered in. The identity infrastructure
that underpins the internet is also the foundation
of the targeted advertising model that finances
online services ranging from search to social
media to email to news to user-provided content.
This business model has driven extraordinary
innovation online, and created high-value
network effects in accordance with Metcalfe’s
Law that the value of a network is proportional
to the square of the number of nodes. Privacy
may be a good that most people are willing
to trade away, and so, on this argument, why
should they be prevented from so doing?
Furthermore, in the current context where far more
information is being created and shared, there
may be reasons to revisit some of the assumptions
underlying the internet’s design. For instance, the
early text-based applications in the internet did not
really cause much of a problem when packets were
indiscriminately sent hither and thither. However,
when the internet is used for synchronous
communication or other time-sensitive
applications, there are limits to acceptable levels of
latency, and certain media, such as video, consume
scarce bandwidth. Using the market to solve issues
of latency and bandwidth implies the development
of property rights to allow infrastructure owners
to make managerial decisions to hold up or
speed up traffic, legitimizing the use of intrusive
technologies such as deep package inspection.
This paper argues that each of these three
responses to the internet’s design, architecture
and governance underpins a particular view of
how it should be run, competing with the original
purist libertarian vision. Hence, there are (at least)
four possible internets. Moreover, each of these
four visions has a powerful set of institutional and
ideological champions, and they can all coexist
— hopefully, but not necessarily, peacefully.
A coda: the internet requires design, standards
and cooperative behaviour. This necessity
implies one final response to the internet, a
human response that elaborate systems tend to
Four Internets: The Geopolitics of Digital Governance 5

invite — subversion. Plain vandalism is a possible response to the complexity and elegance of the internet, and it appears in the form of deliberate and malicious information pollution — trolling being perhaps the most obvious manifestation. However, subversion has an aesthetic of its own, a hacking aesthetic that is pleased to undermine the basic functions or promises of a system, often by using those basic functions against the system itself. Accordingly, ideas such as fake news or the spreading of malware — interventions that would not be possible without the very infrastructure they are there to undermine — are important parts of the subversive’s arsenal. The subversive aesthetic also drives a global position on the internet, originally a dispersed and ad hoc response that manifested itself as cybercrime and hacking, but which in more recent years has itself attracted institutional backers at the level of the nation state.

The Geopolitics of

Internet Ideals

The ideals sketched above are not the only responses to the Silicon Valley ideology of openness, but they are important in 2019 as they all have institutional backing at the level of the nation state or supranational entity. Much of the internet revolves around standards, and an accountable, open and transparent standard-setting process. However, this does not mean that governments are not under pressure to intervene, as either regulators or developers, or via procurement (DeNardis 2014a, 84). Many nations, at least when going through idealistic and optimistic periods (often coinciding with economic growth), have supported open standards, as, for example, India and Brazil in recent years. However, many social effects of the internet, including the spread of social media, the perceived threats to individuals’ (in particular, children’s) psychological well-being, cybercrime, cyberwarfare and a coarsening of public debate, have led some governments to step in more assertively. Above all, the perception that the internet is of necessity a disseminator of liberal and democratic ideals has caused pushback (Morozov 2011). Certain issues, such as net neutrality (see below), or the extent of liability of content platforms or information intermediaries

for the information they carry, fall directly
within governments’ remit to legislate or not.
Governments, therefore, do have power to
shape the internet and to reconfigure the trust
relationships on it, perhaps through what DeNardis
calls the “dark arts” of internet governance
(2014a, 199–221). For example, trusting the
websites we access depends on the maintenance
by Web browsers of lists of trusted certifying
and authenticating authorities. However, such
lists do not solve the problem of online trust,
but rather shift it toward the authorities, which
provide economies of scale in evaluating the
trustworthiness of websites, but which also
create the greater systemic risk of a global rather
than a local model of trust (O’Hara 2004). Such
a system is only as secure as the least common
denominator. A government could compel a
browser-trusted authority to certify an imposter
mail server, for instance, to support surveillance
of its citizens or residents in its territory (DeNardis
2014a, 95). In 2008, the Pakistani government took
down YouTube in Pakistan using the tactic of
requesting Pakistan Telecom to redirect YouTube’s
IP addresses (Hunter 2008). Routing systems
were set up for a smaller and more socially
homogeneous internet, where trust, good faith
and similar aims could be assumed. Of course,
the internet community responds to trust deficits
with improvements in security technology,
but any technical solution lives in some social,
political and economic context as part of a socio-
technical system that is much harder to predict
or control than its technological component.
There are certain types of content that most
governments try to curb, such as child pornography
or pirated intellectual property. There are other
areas, such as political discussion, Holocaust denial
or blasphemy, where (a) only some governments
wish to intervene, and (b) typically they do not
agree on what to censor. However, this does not
mean that they will not try. An important means
for governments to control or censor the content
distributed on the internet is to intervene in
the protocols, the systems or the technology,
as with the Pakistani takedown of YouTube.
Such censorship is not unavoidable — the “dark
Web” often provides technologies to circumvent
such interventions — but it is pretty effective in
stopping messages being disseminated through
audiences whose interest is more casual.

6 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

These powers, however limited, mean that
governments actions are implementing
different conceptions of what the internet
can be. This section reviews the four internets
of most prominence. In addition, this
section considers another vision, not of the
internet per se but of an important rogue
model for understanding the trajectory of
internet governance in 2019 and beyond.

Silicon Valley’s Open Internet

Silicon Valley’s open internet is mainly driven by
the technology. Problems are expected to have
technical solutions primarily, even if there may be
issues about how to implement these. For instance,
WhatsApp is making strides in slowing down the
viral spread of fake news and dangerous rumours
with technical means (which may be easier because
it does not rely on an advertising model; see The
Economist 2018l). With respect to privacy, the most
prevalent view is to see a privacy breach as a
tort (Prosser 1960), requiring the victim to show
evidence of harm.^5 This common law approach
to privacy fits in nicely with the Silicon Valley
credo of “move fast and break things” — innovate
until the innovation is shown to be harmful.
However, not all regulation is bad, on this
view; regulation may be needed to ensure the
unfettered flow of information. Net neutrality
is a signature policy of the Silicon Valley open
internet. It is the principle that internet providers
should not discriminate between different types
of packets of information transmitted over the
internet, to give preferential treatment to some
types over others. Discrimination might happen
for engineering reasons (certain information-
heavy and time-sensitive uses, such as video or
game streaming, might clog up the network),
economic reasons (a mobile operator might
not wish to provide the infrastructure for free
Voice over Internet Protocol [VoIP] services), or
ideological reasons (an operator might wish to
discriminate against child pornography, say, or
the messaging of an opposition political party).
Net neutrality has more of an impact on the last
mile of internet delivery than on global governance.
In countries with sufficient competition between
5 Daniel Solove has written a series of blogs developing a theory about
this. He argues that US courts tend only to see privacy breaches as
harmful if they cause either physical or financial injury, and if the harm
has already happened (ignoring risk of future harm) (Solove 2014).
providers, it is less of an issue, because anyone
who objected to such discrimination could
simply switch to a provider that respected net
neutrality. As an issue, it looms largest in the
United States, where competition is relatively
thin, and where free speech is a highly prominent
shared and constitutionally enshrined value.
Engineers, including Vinton Cerf and Tim Berners-
Lee, have tended to favour net neutrality because
of its positive effects on the network’s efficiency.
Other supporters, however, have been motivated
by business reasons; Google, Amazon and eBay
want as much access to their popular sites as
possible, while companies that offer VoIP services
(such as Microsoft, which owns Skype), and
streaming companies such as Netflix need to avoid
their content being throttled or slowed down.

Brussels’ Bourgeois Internet

Europe’s political attitudes differ from those of
the United States, whose political and public
space are defined by a liberal creed. In Europe,
history plays a much larger role — nation states
have learned, through war, to focus on peace,
prosperity and cohesion. The European Union
was originally posited as an end point to these
integrative processes, and, in cyberspace, it has
taken it upon itself to defend a civilized bourgeois
public space against incivility, taking action, for
example, against disruptors such as Airbnb, which
is blamed for swamping beautiful cities with
tourists ( The Economist 2018k). The European Union’s
Competition Commissioner Margrethe Vestager
has extended the Commission’s anti-trust work
against dominant firms, based on article 102 of
the EU treaty,^6 to pursue American tech giants on
the ground that they might swallow rivals or force
them out of business, leaving consumers with a
poorer standard of service ( The Economist 2017b).^
The bourgeois world rests upon virtuous behaviour,
civility and prudence (McCloskey 2006), and
Western European governments by and large
attempt together with the European Union to
secure this world. Only in such an atmosphere
of trust in government would it be likely that,
for example, Swedes would take to inserting
6 Consolidated Version of the Treaty on the Functioning of the European
Union , 13 December 2007, [2012] OJ, C 326/47, art 102 (entered into
force 26 October 2012), online: <https://eur-lex.europa.eu/legal-
Four Internets: The Geopolitics of Digital Governance 7

microchips in their bodies so enthusiastically ( The Economist 2018n). European thinking on ethics and privacy focuses on dignity, whereas the American tradition looks toward liberty (Whitman 2004), so it is not surprising to find an EU Ethics Advisory Group worrying about the relationship between personhood and personal data, the risks of discrimination as a result of data processing, and the risks of undermining the foundations of democracy (EDPS Ethics Advisory Group 2018).

European courts are regulating the internet increasingly aggressively. To take one prominent example, the Court of Justice of the European Union ruled against Google Spain in 2014 in a case brought by a man who wanted outdated information about him removed from Google’s search results.^7 The original decision was a compromise, and a controversial one, although welcomed by many commentators, including the present authors (O’Hara 2015; O’Hara and Shadbolt 2015; O’Hara, Shadbolt and Hall 2016), as allowing the European Union to police its own jurisdiction without imposing its own restrictive view of privacy upon the world. However, since the judgment, the French data protection regulator CNIL (Commission Nationale de l’Informatique et des Libertés) has tried to push back against searches for EU citizens in any jurisdiction, and the General Data Protection Regulation (GDPR) of 2018 has enshrined that universalism into EU law, even switching the emphasis from delisting to erasure (Politou, Alepis and Patsakis 2018).

Many suspect it will be harder to innovate in Brussels’ bourgeois internet, thanks to a preference for incumbents and distaste for disruptive newcomers. For example, the GDPR is perceived as a threat to the model of free services for surveillance ( The Economist 2018c). The GDPR is a paradigm case of the European Union’s drive to a bourgeois level of safety. In contrast to American law, it covers every kind of data processing, whether shown to be harmful or not, and tries to anticipate and minimize risk (although it has been argued that the box-ticking mentality it has promoted is in practice no more protective of privacy than the tort-based approach of the United States; see Bamberger and Mulligan [2015]). Yet, the GDPR remains a source of advantage for the

7 Google Spain SL and Google Inc v Agencia Española de Protección de Datos (AEPD) and Mario Costeja González (13 May 2014), Doc C-131/12, ECLI:EU:C:2014:317 (CJEU).

European Union — it is a leader in data protection
because it is too large a market to ignore. It is also
totemic: “This new data protection ecosystem
stems from the strong roots of another kind of
ecosystem: the European project itself, that of
unifying the values drawn from a shared historical
experience with a process of industrial, political,
economic and social integration of States, in order
to sustain peace, collaboration, social welfare and
economic development” (EDPS Ethics Advisory
Group 2018, 6). The jury is out; the GDPR has
certainly been influential worldwide. However,
it may handicap Europe in the development of
artificial intelligence (AI). Where China and the
United States are each large centralized markets,
enabling the gathering of giant quantities of
data to fuel their algorithms, Europe is more
fragmented, both in terms of markets and in
terms of the dominant tech companies, and this
decentralization is exacerbated by the GDPR’s
stern regulation of data sharing (China’s data
advantage is discussed in the next section).
Privacy is not, of course, the only area where the
European Union’s instinct is skeptical of market
forces, which are sometimes perceived as too
disruptive, creating social costs, and sometimes
perceived as producing an incoherent or inefficient
internet where private gain crowds out public
gain. A satisfactory set of arrangements is simply
inconceivable without a regulator. For instance,
the European Union’s update of its copyright laws^8
has attracted opprobrium because of its aggressive
stance on copyright breaches ( The Economist
2018u). Characteristic of the European Union’s
attitude toward technology firms is its assumption
that complaints about regulation threatening
the freewheeling, entrepreneurial internet are
exaggerated. Article 13 of the new copyright
law compels internet firms to work closely
with copyright holders to bring down copyright
materials as soon as possible, which (given the
imprecise nature of copyright identification
algorithms) is likely to result in overzealous
policing. Article 11 requires aggregators to obtain
a licence from publishers if they display excerpts
from content. A similar rule introduced in Spain
in 2014 led Google to withdraw its aggregation
service from there; the bet underlying article 11
is that Google could not afford to do the same
8 The Directive of the European Parliament and of the Council on copyright
in the Digital Single Market, at http://www.consilium.europa.eu/media/35373/

8 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

across the whole of Europe. The European mindset
is that reasonable behaviour is unlikely (if not
impossible) in the absence of rules: one study for
the European Union about the interconnection
of the internet’s autonomous systems concluded:
“A recurrent theme in the discussion of IP
interconnection is whether network operators
will be motivated to interconnect (on reasonable
terms) in the absence of a regulatory obligation”
(quoted in DeNardis 2014a, 130). Meanwhile, some
agencies are simply acting to try to influence, as,
for example, the United Kingdom’s Government
Communications Headquarters helping UK
cyber security firms ( The Economist 2018o).
Government, on this model, is the primary locus
of trust. It is doubtful whether this proposition
would be supported (or supportable) on any of
the other internets described in this paper.

Beijing’s Authoritarian Internet

China’s importance not only to the world economy,
but also to the internet, has grown remarkably
in recent years, so that over half the country
is connected to the internet, and over half of
internet users are in China. Even if one is skeptical
of Robert Kaplan’s claim (2018) that the classic
geography of the Eurasian empires has returned,
the new assertiveness of China has coincided
with a shift in European and Asian geopolitics,
which has led to a diminution of the constraints
of behaviour on China (and also on Russia).
The internet, for China, has been a boon for
surveillance. Technology, for example, is used to
monitor restive populations such as that of Xinjiang
province ( The Economist 2018d). Protections against
surveillance are being eroded across the globe, as
the technology becomes easier to apply and people
are more willing to behave in ways that make
them easy to watch, such as social networking.
However, the trend is particularly strong in China.
The Chinese model is based on the promotion of
its own tech giants, Baidu, Tencent and Alibaba.
These are private companies, astonishingly
successful in their own right, but operating within
a tightly controlled environment in which the
ruling Communist Party is the dominant player.
However, China also has an increasing presence on
international bodies; for instance, it currently holds
the chair of the International Telecommunication
Union, which is pushing for standards that will aid
government micromanagement of the internet.
China itself has begun to invest heavily in
technology using venture capital models ( The
Economist 2018j). Alibaba and Tencent are among
the largest of China’s venture capital investors and
are shaping the start-up world in that nation ( The
Economist 2018p). Beyond that, Chinese companies
have invested billions of dollars of venture capital
in US start-ups, despite pushback from President
Trump’s administration and the European Union
( The Economist 2018m; 2018q). Much of this activity
is helped by the specific ways in which Chinese
firms have adapted to the Chinese business
environment, which is characterized by shaky
rule of law, massive consumer scale, extremely
changeable demand, cutthroat competition and
proximity to an efficient low-cost manufacturing
hub. Whereas US firms have developed to take
advantage of their own more stable and business-
friendly environment, with high breadth of
ownership and relatively transparent management,
Chinese firms are often closely associated with
a celebrity boss/owner with majority control
(unlike even Mark Zuckerberg or Steve Jobs),
and display highly opportunistic behaviour,
expanding quickly into new markets (thereby
resembling the sprawling conglomerates of old)
( The Economist 2018g). Kai-Fu Lee has argued that
Chinese companies are hungrier, less complacent,
more vigorous, more eager for competition, and
less constrained by mission statements and core
values than their US counterparts (Lee 2018).
Furthermore, according to the same author, the
age of the massive AI breakthroughs, where the
United States has been a leader, is being superseded
by an age of implementation, of applying and
adapting the algorithms to the dull problems of
everyday life. Here, China has the advantage, both
in terms of the national skillset, and in terms of
the numbers of scientists it can deploy (ibid.).
Another growing source of advantage for China
is its trove of data, the raw material of AI (ibid.).
China’s internet economy generates far more data
than any other, partly because of its size and partly
because much Chinese commerce has moved on
from cash to electronic payments. The social media
app WeChat has become dominant in China, not
only for communication with friends, family and
work colleagues but also for mobile payments.
Expatriate Chinese are increasingly using WeChat
and it has started to spread throughout the West as
a result. All the data is stored in China and therefore
accessible to the Chinese government. Furthermore,
unhindered by data protection regulation or
Four Internets: The Geopolitics of Digital Governance 9

noticeable public demand for privacy, data is gathered from many other sources, including closed circuit television. This data is immensely important to Chinese science but also augmented by various schemes in which Chinese citizens rate each other as citizens on social networks. China hopes to lead in AI and has made advances in areas such as face recognition and autonomous vehicles. Its less-developed status helps as well, in terms of social and industrial adaptability; whereas the United States is restricting the use of self-driving cars and worrying about pedestrian deaths ( The Economist 2018b), China is building a city to accommodate them (Lee 2018).

Beyond its borders, China’s influence on American firms is growing. In 2018, it forced Apple to transfer its iCloud data about Chinese users to a Chinese data centre ( The Economist 2018r). Of course, this kind of nationalism is common across the world, including in the European Union, but it does mean that the government can certainly get hold of this valuable data more easily. Business in the lucrative Chinese market will have to be done on Chinese terms. In 2010, Google quit China in order to avoid having to censor search results. At the time of writing, it is reported that Google is testing a mobile search app called “Dragonfly,” which would filter websites blocked by China’s “Great Firewall,” and provide instead a notice that some results might have been removed. If it goes ahead, it would have to compete with Baidu, which carries out 75 percent of searches in China, and which has cemented its dominance by ensuring that its own apps are pre-installed on Chinese smartphones (ibid.).^

In 2013, President Xi Jinping unveiled an infrastructure and trade initiative, entitled the Silk Road Economic Belt and the Twenty-first- Century Maritime Silk Road, often called “The Belt and Road Initiative.” This aims to link together the Eurasian world with connectivity and cooperative ventures, as a route for future Chinese (and other) trade, by developing infrastructure across Asia, Europe and Africa. The authoritarian internet could well become part of this project, leading to a Belt, Road and Information Superhighway Initiative, comprising the technological areas where China sees potential advantage, including AI, big data, quantum computing and cloud storage. The city of Xi’an in Shaanxi province, a bastion of the original medieval Silk Road, has already positioned itself as a tech centre ( The Economist 2018e).^

Such an internet might easily be supported by
poorer countries for which the internet has proved
problematic — for instance, countries including
Mauretania, Algeria, Uzbekistan, Iraq and Ethiopia
have been forced to turn the internet off during
school exam time, because of the prevalence of
cheating ( The Economist 2018h). While Chinese
companies have been increasingly targeted by
nationalists in the United States, and major US
firms apart from Amazon and Apple are pretty
well barred from China, the major Chinese and
American firms compete in other markets, such
as Brazil, Indonesia, India and Africa. In January
2019, the planet is on course to achieve a figure
of 50 percent of its population connected to the
internet, but with much of the remaining 50
percent in rural China, India and Africa. China has
a considerable financial influence in Africa and will
seek to influence the governance of the internet
there. It may do this under the radar; while US firms
tend to transplant their usual services into the new
markets under their own names, tweaking where
necessary, Chinese firms have a somewhat more
covert strategy of buying stakes in promising start-
ups (as they have even in the United States before
getting pushback from the Trump administration)
— 2017 saw US$5 billion invested in Indian start-
ups by Chinese tech firms ( The Economist 2018i).^

DC’s Commercial Internet

The characteristics of what might be called the
“DC commercial internet” — the vision of the
commercial internet as espoused by leaders in the
US Capitol — are similar to those of the Silicon
Valley open internet — and indeed, commercial
and technology interests have always cooperated
strongly through the internet’s history, helped
by their geographical concentration in the
same nation. However, the United States is now
polarized to an unprecedented degree, and the
champions of the DC model, in particular the
Republicans, notably President Trump, are at
loggerheads (over a tremendous number of issues)
with the champions of the Silicon Valley model,
in particular the Democrats and Barack Obama,
whose White House hosted a number of present
and former technologists. Most prominently, the
Federal Communications Commission (FCC) voted
in December 2017 by three to two to repeal its
commitment to net neutrality that it had brought
in under the previous administration in 2015. The
Star Wars actor Mark Hamill criticized the FCC
for siding with large corporations against the

10 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

individual; Senator Ted Cruz replied that Darth
Vader would have approved of regulating the
internet ( The Economist 2017c). The head of the FCC,
Ajit Pai (appointed to the FCC by President Obama,
but elevated to its chair by President Trump),
claims to be a supporter of net neutrality but argues
that federal regulation will suppress innovation,
and that net neutrality ought to be a contractual
matter between ISPs and their customers, in their
terms and conditions ( The Economist 2017a).^
The roots of the Silicon Valley/DC split lie in the
collective action problem that affects internet
operators (DeNardis 2014a). These operators
compete with each other for their customers, but
on the other hand, their cooperation in connecting
their networks with each other, using standard
protocols and handling their competitors’ traffic,
makes the internet the internet, rather than a series
of disconnected or weakly connected islands. This
creates a tension between what we might think of
as the public good of a seamless internet, and the
private interests of these operators. The tension
has led to much creation and innovation, but the
line between the public and the private good can
shift. Silicon Valley’s open internet focuses on the
public, while DC’s commercial internet leverages
the interests of private actors, on the argument
that large profits show that public interests are
indeed being served by these self-interested actors.
For example, in contrast to the European Union’s
approach under Commissioner Vestager (see
above), US trustbusters are more tolerant of the
monopolistic tendencies of the industry, following
an argument of Joseph Schumpeter that the
promise of monopoly profits can be an important
driver of innovation and customer service
(Schumpeter 2010, 76–92). Having said that, even
the Silicon Valley firms can be torn. Facebook, for
example, is a prime builder of the walled gardens
that Jonathan Zittrain (2008) railed against, while
the tech giants are so keen to buy start-ups that
they are threatening the start-up culture for which
Silicon Valley is famous ( The Economist 2018f; 2018t).^
This dilemma is exacerbated by its being located
in the United States, where the extent and limits
of free speech are a matter of major constitutional
interest. The First Amendment forbids the state
to curb free speech, but jurisprudence has led
to divergent interpretations. The affirmative
interpretation, which held sway during much
of the twentieth century, holds that the state
is justified in intervening in public spaces for
expression (even ones that are privately owned,
such as telecommunications and internet
spaces), to support the societal goal of facilitating
expression of a multiplicity of viewpoints and,
conversely, restricting the rights of the owners
of these spaces to censor or limit the messages
they carry. The negative interpretation is that
the First Amendment forbids the state from
intervening in such spaces, as to do so would
restrict the free speech rights of the owners to
determine what voices are heard in their spaces;
this interpretation has been the majority view of
the Supreme Court since the 1980s. In short, does
the state have a positive duty to make sure speech
is promoted, even on private property, or does
the First Amendment’s scope only cover publicly
administered spaces, so that private property
owners’ rights are unaffected by it (Nunziato 2009)?
The positive interpretation favours net neutrality
and Silicon Valley openness, while the negative
interpretation (at the time of writing in the legal
ascendant) favours private property interests. It is
a parochial argument to the rest of us, but the way
it plays out will affect the internet as a whole.
There are engineering arguments for limited
traffic discrimination, such as to manage the
network and to ensure that quality of service is
maintained for all — for example, during busy
periods, it might be acceptable to slow down
content that is not so time-critical, such as email.
But most arguments against net neutrality have
strong business reasons. Internet providers are
the organizations that would have to obey any net
neutrality law, such as the regulations brought
in by the FCC in 2015 under Obama, and they
are generally opposed, preferring not to have
constraints on their network management. Neo-
liberal free-market thinkers are also opposed, not
only because they generally oppose government
regulation on ideological grounds but also because
they support free market solutions to problems
based on freedom to exploit property rights (the
providers are seen, on this view, as owners of
the network, and so should be free to manage
them as they see fit). If anyone is treated unfairly,
then they should have recourse to a private legal
challenge, rather than protection via regulation,
so that regulation would happen “organically” via
common law. We see here a clash between two
types of liberty supported under liberalism, as
described by Isaiah Berlin (2002): “freedom from”
(in this case, censorship) versus “freedom to” (in
Four Internets: The Geopolitics of Digital Governance 11

this case, manage one’s private property, that is, the internet infrastructure owned by providers).

More widely, this property-based model threatens the interoperability that was a fundamental principle of the internet and, subsequently, the Web — Berners-Lee in his 2018 Turing Lecture^9 argued that the universality of identifiers for online resources was key for the added value of the Web. As early as 2008, Zittrain sounded an alarm about what he called non-generative models of the internet, which created walled gardens and undermined innovation (Zittrain 2008). Since Zittrain wrote, the extraordinary growth of social networking has built the walls around the gardens still higher, while arguably making the gardens prettier and more habitable.

In particular, social networking sites (SNSs) bypass some of the internet’s interoperability mechanisms. They do not particularly support cross-platform compatibility (so that interacting between two SNSs is not as simple as, say, sending an email from Gmail to an .edu address). Personal data is not portable between sites, although the GDPR is attempting to change this. Search is restricted. Resources are not identified or located by universal formalisms (DeNardis 2014b). As Berners-Lee wrote in 2010, “connections among data exist only within a site. So the more you enter, the more you become locked in. Your [SNS] becomes…a closed silo of content…The more this kind of architecture gains widespread use, the more the Web becomes fragmented, and the less we enjoy a single, universal information space” (quoted in DeNardis 2014a, 241). Zittrain and Berners-Lee defend the Silicon Valley open internet, but the DC commercial response is that SNSs provide services that people actually wish to access, in large numbers, and that the only responsibilities SNS owners have are to their customers, assuming that they do not interfere with the running of the internet as a whole. As with other types of property, if someone wishes to build a wall around their garden, they should be allowed to do so as long as they cause no harms elsewhere. They should be the best judge of the value to be obtained from their property. The single, universal information space that Berners- Lee advocates cannot and should not be imposed, on this view, against the will of someone to monetize their intellectual property via restriction.

9 See https://amturing.acm.org/vp/berners-lee_8087960.cfm.

Addendum: Moscow’s

Spoiler Model

As noted above, geopolitical shifts have led to
a lessening of the constraints on Russia and a
reassertion of the imperial geography of the past
(Kaplan 2018). Russia under President Vladimir
Putin has exploited this to engineer an ideological
space opposed to the West, based on a mystical
mélange of nationalism and destiny, ressentiment
and victimhood, power and calculation, cynicism
and conspiracy theories (Snyder 2018). Given
this vision, the decentralized internet, with no
institutionalized editing or fact-checking, has been
an ally. Indeed, the polarization of politics in the
West, notably in the United States but also in the
European Union, has provided the opportunity
to import the uncertainties and obfuscations
routine in Russian politics into Western politics,
by cheaply importing narratives, arguments
and conspiracies using the power of bots. Much
of this has been revealed by Robert Mueller’s
inquiry into Russian interference in the 2016 US
presidential election ( The Economist 2018a).^
There are several other instances of this, which
appear strategically inexplicable except as a means
of sowing division and mistrust. For instance, David
A. Broniatowski et al. (2018) report that Russian
bots and trolls regularly tweet about vaccination in
divisive terms, linking the issues to controversies
in American politics. The tweets are both pro- and
anti-vaccination, but the purpose appears to be less
to establish a position as to create, by the volume
of tweets, the impression of strong and partisan
debate, and to recruit partisan campaigners by
associating vaccination with the several other
wedge issues in America’s dysfunctional politics.
This is not just a Russian tactic (although the term
“disinformation” was indeed originally a Russian
term, coined during the Stalin era). No doubt
all nations indulge in deliberately propagating
falsehood. However, disinformation is a particularly
potent weapon against the West, where speech is
freer (and it is easier to spread ideas), and where
controlling the public sphere is seen as rather alien.
A recent report from the Oxford Internet Institute
argued that “computational propaganda is now
one of the most powerful tools against democracy”
(Woolley and Howard 2017, 7) and found evidence
that, for instance, 45 percent of Twitter activity
in Russia was automated for the creation of
disinformation (ibid., 4), and that political debate in
Germany, the United States, Poland, Brazil, Ukraine

12 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

and Taiwan is also compromised (ibid.). In August
2018, Facebook and Twitter shut down hundreds of
accounts accused of spreading disinformation not
only from Russia, but also from Iran (Timberg 2018).

Discussion: Four Internets

and a Free Rider

These five visions of the internet do not, and
probably could not, exist in their pure forms, still
less be so neatly ascribed to particular regimes.
They are caricatured here to make the main points:
the homogeneity of the internet cannot be assumed
(Global Commission on Internet Governance 2016),
and scenarios about what is sometimes called its
Balkanization (creating “the Splinternet”) cannot
be ruled out. Neither are these the only internets
that could evolve — the four (plus one) could
become five, or six, or seven or more. There could
be a developing world internet, or a feminist
internet, or an Islamic internet, or a caring internet,
or an internet of cyborgs, if the appropriate
ethical vision found a technological realization
and sufficiently powerful institutional backing.
Many commentators have drawn the conclusion
that this is a straight fight between China and
the United States.^10 This notion underestimates
the breadth of dispute between conflicting
visions (not least within the United States
itself ). However, it is important to understand
as well that these models do (at the moment)
coexist in uneasy tension, and that (so far) all are
perceived to have some value by most actors.
Russia is singled out as the spoiler, free riding
on the efforts of others to produce a valuable
information space. Of course, very many
nations, including the United States, indulge in
disinformation. The actions of the United States
(under both Obama and Trump) in indicting cyber
spies and cyber warriors from China, Russia, Iran
and elsewhere have reportedly concerned members
of its National Security Agency, who themselves
fear being prosecuted outside the United States for
similar crimes ( The Economist 2018v). Meanwhile,
10 For example, Eric Schmidt (quoted in Kolodny 2018) and Lee, in his book
AI Superpowers: China, Silicon Valley and the New World Order (2018).
although the Russians and others are happy to troll
the internet, they do require a functioning internet
to troll, so they have no incentive to undermine it
totally (both the honest and the dishonest benefit
from general honesty; compare Nyberg 1997;
Iñiguez et al. 2014). However, the acceptability
of dishonesty is likely to increase if the system
as a whole is perceived as unfair, providing
spoilers with incentives to highlight lapses in the
standards of other nations with accusations of
hypocrisy (compare, for example, Zhang 2008).
Similarly, the Chinese authoritarian model appeals
to its government, which is quick to close down
conversation in its lively microblogging media.
However, it also values the openness that leads
to the publication of dissent, which it uses as
an early warning of problems with illegal land
appropriations, pollution, corruption, poor food
and air quality, and other issues. Conversely,
the authoritarian internet will appeal to any
government, however democratic, that takes
responsibility for social problems (such as obesity
or climate change) and would rather impose a
paternalistic solution than allow one to emerge
from an autonomous citizenry; the kind of soft
paternalism known as the “nudge” philosophy
is one means of leveraging large quantities of
data within an internet environment in which
choices are carefully closed down (Thaler and
Sunstein 2008). India, for example, eschews the
full Chinese authoritarian suite, but nevertheless
has access to large quantities of social media and
banking data that are highly linkable through
its Aadhaar digital biometric identity scheme.
In the United States, as emphasized above, the
breakdown of political consensus has made the
distinction between the Silicon Valley open internet
and the DC commercial internet far sharper than it
traditionally has been (one of the last acts in office
of President Trump’s former Attorney General Jeff
Sessions was to sue the State of California for its
decision to restore net neutrality regulation against
the FCC’s own reversal [ The Guardian 2018]), but
until fairly recently the two visions managed to rub
along reasonably well, with businesses switching
their evangelizing between openness and property/
markets opportunistically as their situations
demanded. Meanwhile, some of the tech giants
are recruiting prominent European politicians to
explain their positions to fellow Eurocrats, such
as Facebook’s appointment of the former leader of
Britain’s Liberal Democrats, Nick Clegg, as its head
Four Internets: The Geopolitics of Digital Governance 13

of global affairs (Clegg 2018). Such cross-fertilization may also result in bringing the Eurocrats closer to the Americans; Margrethe Vestager’s decision in 2017 to order Apple to pay back-taxes to the Irish government (that did not want the money) was criticized by one of her predecessors, Neelie Kroes, who had been appointed to Uber’s Public Policy Advisory Board in 2016 ( The Economist 2017b).^

Even Tim Berners-Lee, a consistent apostle of openness, has a vision of the Web that looks much closer to Brussels’ bourgeois internet than Silicon Valley’s open one, in which polite conversation is not drowned out by the roughhouse — consistent with the Web’s birth as a means of disseminating scientific research (Berners-Lee 2018). The initiatives he has championed — ranging from the Web We Want,^11 a project of the World Wide Web Foundation, to a “Magna Carta for the Web” (Kiss 2014; Sample 2018), to the Solid platform, which is intended to “re-decentralize” the Web guided by the principle of “personal empowerment through data”^12 — aim to promote human rights, privacy, anti-discrimination and trolling, and bear a closer resemblance to the European Commission’s vision than to John Perry Barlow’s. The Solid vision sees individuals curating their own data responsibly and managing read/write permissions via “PODs” — personal online data stores — thereby meeting one of Berners-Lee’s own worries about the Web (that we have lost control of our personal data), but maybe not dealing with some of the by-products of openness, specifically the spread of misinformation and the lack of transparency (Berners-Lee 2017). The Global Commission on Internet Governance (2016) adopts a similar position of combining openness with a respectful environment.

Hence these models (and the spoilers that undermine them) are likely to coexist even within individual organizations and governments. Nevertheless, clear preferences exist for certain models, and these contribute to the tensions in global internet governance.

11 See https://webwewant.org/.

12 See https://solid.mit.edu/.


In 2002, when the world seemed unipolar under
a benign if stern American hegemony, and the
recent terrorist attacks in New York had created an
imperative to reassert American moral ascendency,
President George W. Bush described an “Axis
of Evil.” In today’s very different world, we can
discern a somewhat scarier “Axis of Incivility,”
of nation states jostling for narrow advantage,
with a view of international relations, including
economic relations, as zero sum. Unlike the Axis of
Evil, which reflected US foreign policy concerns,
the Axis of Incivility has at its foundations the
three major superpowers, the United States, China
and Russia, each of which in its different ways at
the time of writing pursues aggressive nationalist
policy goals while showing impatience with
due process both internally and internationally.
Many other nations, including Egypt, Hungary,
India, Iran, Israel, the Philippines, Poland, Saudi
Arabia and Turkey, are following this lead.
In such a world, it is inconceivable that these
competing visions of the internet will not
become entangled in the drive for international
recognition, power and coalition-building.
Neither the benefits of cooperation and openness,
nor those of privacy and bourgeois stability,
are likely to cut much ice with rational actors
with such a mindset. Hence, the competition
to establish which, if any, of the four internets
will prevail (however temporarily) is likely to be
strong, and not always focused on win-wins.

14 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

Works Cited

Bamberger, Kenneth A. and Deirdre K. Mulligan.
  1. Privacy on the Ground: Driving Corporate Behavior in the United States and Europe. Cambridge, MA: MIT Press.
Barlow, John Perry. 1996. “A Declaration of
the Independence of Cyberspace.”
Davos, Switzerland, February 8.
Benkler, Yochai. 2006. The Wealth of Networks:
How Social Production Transforms Markets and
Freedom. New Haven, CT: Yale University Press.
Berlin, Isaiah. 2002. “Two concepts of liberty.”
In Liberty , edited by Henry Hardy, 166–217.
Oxford, UK: Oxford University Press.
Berners-Lee, Tim. 2017. “I invented the web. Here are
three things we need to change to save it.” The
Guardian , March 12. http://www.theguardian.com/
———. 2018. “One Small Step for the Web...”
Inrupt.com, September 28. http://www.inrupt.com/
Berners-Lee, Tim, Wendy Hall, James A.
Hendler, Kieron O’Hara, Nigel Shadbolt and
Daniel J. Weitzner. 2006. “A Framework
for Web Science.” Foundations and
Trends in Web Science 1 (1): 1–134.
Blum, Andrew. 2012. Tubes: Behind the Scenes
at the Internet. London, UK: Viking.
Broniatowski, David A., Amelia M. Jamison, SiHua
Qi, Lulwah AlKulaib, Tao Chen, Adrian Benton,
Sandra C. Quinn and Mark Dredze. 2018.
“Weaponized health communication: Twitter
bots and Russian trolls amplify the vaccine
debate.” American Journal of Public Health 108
(10): 1378–84. https://ajph.aphapublications.
Clegg, Nick. 2018. “I’m joining Facebook to build
bridges between politics and tech.” The
Guardian , October 19. http://www.theguardian.com/
DeNardis, Laura. 2014a. The Global War
for Internet Governance. New Haven,
CT: Yale University Press.
———. 2014b. “The social media challenge to
Internet governance.” In Society and the
Internet: How Networks of Information and
Communication Are Changing Our Lives , edited
by Mark Graham and William H. Dutton,
348–59. Oxford, UK: Oxford University Press.
EDPS Ethics Advisory Group. 2018. Towards a
Digital Ethics. European Data Protection
Supervisor. https://edps.europa.eu/sites/edp/
Global Commission on Internet Governance.
  1. One Internet. Waterloo, ON: CIGI. http://www.cigionline.org/publications/one-internet.
Hunter, Philip. 2008. “Pakistan YouTube block
exposes fundamental Internet security
weakness: Concern that Pakistani action
affected YouTube access elsewhere in world.”
Computer Fraud and Security 4: 10–11.
Iñiguez, Gerardo, Tzipe Govezensky, Robin
Dunbar, Kimmo Kaski and Rafael A.
Barrio. 2014. “Effects of deception in
social networks.” Proceedings of the Royal
Society B: Biological Sciences 281 (1790).
Kaplan, Robert D. 2018. The Return of Marco
Polo’s World: War, Strategy, and American
Interests in the Twenty-First Century.
New York, NY: Random House.
Kiss, Jemima. 2014. “An online Magna Carta:
Berners-Lee calls for bill of rights for web.”
The Guardian , March 12. http://www.theguardian.com/
Kolodny, Lora. 2018. “Former Google CEO predicts
the internet will split in two — and one part
will be led by China.” CNBC Tech, September 20.
Lee, Kai-Fu. 2018. AI Superpowers: China, Silicon
Valley and the New World Order. New
York, NY: Houghton Mifflin Harcourt.
Four Internets: The Geopolitics of Digital Governance 15

McCloskey, Deirdre N. 2006. The Bourgeois Virtues: Ethics for an Age of Commerce. Chicago, IL: University of Chicago Press.

Morozov, Evgeny. 2011. The Net Delusion: How Not to Liberate the World. London, UK: Allen Lane.

Nunziato, Dawn C. 2009. Virtual Freedom: Net Neutrality and Free Speech in the Internet Age. Stanford, CA: Stanford University Press.

Nyberg, Sten. 1997. “The honest society: stability and policy considerations.” Journal of Public Economics 64 (1): 83–99. https://doi.org/ 10.1016/S0047-2727(96)01608-8.

O’Hara, Kieron. 2004. Trust: From Socrates to Spin. Duxford, UK: Icon Books.

———. 2009. “Web engineering in the Chinese context: ‘Let a hundred flowers bloom, a hundred schools of thought contend.’” In China’s Information and Communications Technology Revolution: Social changes and state responses , edited by Xiaoling Zhang and Yongnian Zheng, 121–35. Abingdon, UK: Routledge.

———. 2015. “The Right to Be Forgotten: The Good, the Bad, and the Ugly.” IEEE Internet Computing 19 (4): 73–79. http://doi. ieeecomputersociety.org/10.1109/MIC.2015.88.

———. 2018. “The contradictions of digital modernity.” AI and Society , April 25. https://doi.org/10.1007/s00146-018-0843-7.

O’Hara, Kieron, Noshir S. Contractor, Wendy Hall, James A. Hendler and Nigel Shadbolt.

  1. “Web Science: understanding the emergence of macro-level features on the World Wide Web.” Foundations and Trends in Web Science 4 (2-3): 1–165.

O’Hara, Kieron and Wendy Hall. 2013. “Web Science.” In The Oxford Handbook of Internet Studies , edited by William H. Dutton, 48–68. Oxford, UK: Oxford University Press.

O’Hara, Kieron and Nigel Shadbolt. 2015. “The right to be forgotten: its potential role in a coherent privacy regime.” European Data Protection Law Review 1 (3): 178–89. https://doi.org/10.21552/EDPL/2015/3/5.

O’Hara, Kieron, Nigel Shadbolt and Wendy Hall.
  1. A Pragmatic Approach to the Right to Be Forgotten. Global Commission on Internet Governance Paper Series No. 26. Waterloo, ON: CIGI. http://www.cigionline.org/publications/ pragmatic-approach-right-be-forgotten.
Politou, Eugenia, Efthimios Alepis and Constantinos
Patsakis. 2018. “Forgetting personal data and
revoking consent under the GDPR: challenges
and proposed solutions.” Journal of Cybersecurity.
Prosser, William L. 1960. “Privacy.”
California Law Review 48: 383–423.
Sample, Ian. 2018. “Tim Berners-Lee launches
campaign to save the web from abuse.” The
Guardian , November 4. http://www.theguardian.com/
Schumpeter, Joseph. 2010. Capitalism, Socialism and
Democracy. Abingdon, UK: Routledge Classics.
Snyder, Timothy. 2018. The Road to Unfreedom.
London, UK: Bodley Head.
Solove, Daniel. 2014. “Privacy and Data Security
Violation: What’s the Harm?” TeachPrivacy
(blog), July 2. https://teachprivacy.com/
Thaler, Richard H. and Cass R. Sunstein.
  1. Nudge: Improving Decisions About Health, Wealth, and Happiness. New Haven, CT: Yale University Press.
The Economist. 2017a. “Another debate about net
neutrality in America.” The Economist , April 22.
———. 2017b. “Is Margrethe Vestager
championing consumers or her political
career?” The Economist , September 14.
———. 2017c. “A vote on ‘net neutrality’ has
intensified a battle over the internet’s
future.” The Economist , December 23.
———. 2018a. “Russian disinformation distorts
American and European democracy.”
The Economist , February 24.
———. 2018b. “A pedestrian has been killed by a
self-driving car.” The Economist , March 24.

16 CIGI Papers No. 206 — December 2018 • Kieron O’Hara and Wendy Hall

———. 2018c. “Who will be the main
loser from Europe’s new data-privacy
law?” The Economist , May 26.
———. 2018d. “China has turned Xinjiang into a
police state like no other.” The Economist , June 2.
———. 2018e. “China talks of building a ‘digital
Silk Road.’” The Economist , June 2.
———. 2018f. “American tech giants are making
life tough for startups.” The Economist , June 2.
———. 2018g. “Xiaomi’s forthcoming IPO
shows how the rules of business are
changing.” The Economist , June 9.
———. 2018h. “Why some countries are
turning off the internet on exam
days.” The Economist , July 7.
———. 2018i. “Chinese and US tech giants go at it
in emerging markets.” The Economist , July 7.
———. 2018j. “China’s new $15bn tech
fund emulates SoftBank’s Vision
Fund.” The Economist , July 14.
———. 2018k. “The backlash against
Airbnb.” The Economist , July 21.
———. 2018l. “WhatsApp suggests a cure
for virality.” The Economist , July 28.
———. 2018m. “America and the EU
are both toughening up on foreign
capital.” The Economist , July 28.
———. 2018n. “Why Swedes are inserting
microchips into their bodies.”
The Economist , August 4.
———. 2018o. “Britain’s spies get
entrepreneurial.” The Economist , August 4.
———. 2018p. “Alibaba and Tencent have
become China’s most formidable
investors.” The Economist , August 4.
———. 2018q. “Silicon Valley gets queasy about
Chinese money.” The Economist , August 11.
———. 2018r. “Plans for a return to China has
many up in arms.” The Economist , August 25.
———. 2018s. “Why startups are leaving Silicon
Valley.” The Economist , September 1.
———. 2018t. “Silicon Valley is changing, and
its lead over other tech hubs narrowing.”
The Economist , September 1.
———. 2018u. “A controversial new copyright
law moves a step closer to approval.”
The Economist , September 15.
———. 2018v. “America’s government is
putting foreign cyber-spies in the
dock.” The Economist , September 15.
The Guardian. 2018. “US justice department sues
California over new net neutrality law.” The
Guardian , October 1. http://www.theguardian.com/
Timberg, Craig. 2018. “Facebook, Twitter
remove hundreds of disinformation pages
created by Russia, Iran.” The Sydney Morning
Herald , August 22. http://www.smh.com.au/
Whitman, James Q. 2004. “The two Western
cultures of privacy: dignity versus
liberty.” Yale Law Journal 113: 1151–221.
Woolley, Samuel C. and Philip N. Howard. 2017.
Computational Propaganda Worldwide: Executive
Summary. Working Paper 2017.11. Oxford, UK:
Oxford Internet Institute. http://comprop.oii.
Working Group on Internet Governance. 2005.
Report from the Working Group on Internet
Governance. Document WSIS-II/PC-3/DOC/5-E,
August 3. http://www.itu.int/net/wsis/docs2/pc3/off5.pdf.
Zhang, Yue. 2008. “The effects of perceived
fairness and communication on honesty
and collusion in a multi-agent setting.”
Accounting Review 83 (4): 1125–46. https://
Zittrain, Jonathan. 2008. The Future of
the Internet: And How to Stop It. New
Haven, CT: Yale University Press.

About CIGI

We are the Centre for International Governance Innovation: an independent, non-partisan think tank with an objective and uniquely global perspective. Our research, opinions and public voice make a difference in today’s world by bringing clarity and innovative thinking to global policy making. By working across disciplines and in partnership with the best peers and experts, we are the benchmark for influential research and trusted analysis.

Our research programs focus on governance of the global economy, global security and politics, and international law in collaboration with a range of strategic partners and support from the Government of Canada, the Government of Ontario, as well as founder Jim Balsillie.

À propos du CIGI

Au Centre pour l’innovation dans la gouvernance internationale (CIGI), nous formons un groupe de réflexion indépendant et non partisan doté d’un point de vue objectif et unique de portée mondiale. Nos recherches, nos avis et nos interventions publiques ont des effets réels sur le monde d’aujourd’hui car ils apportent de la clarté et une réflexion novatrice pour l’élaboration des politiques à l’échelle internationale. En raison des travaux accomplis en collaboration et en partenariat avec des pairs et des spécialistes interdisciplinaires des plus compétents, nous sommes devenus une référence grâce à l’influence de nos recherches et à la fiabilité de nos analyses.

Nos programmes de recherche ont trait à la gouvernance dans les domaines suivants : l’économie mondiale, la sécurité et les politiques mondiales, et le droit international, et nous les exécutons avec la collaboration de nombreux partenaires stratégiques et le soutien des gouvernements du Canada et de l’Ontario ainsi que du fondateur du CIGI, Jim Balsillie.

67 Erb Street West Waterloo, ON, Canada N2L 6C2 http://www.cigionline.org