ArticlePDF Available

Network Neutrality: what is internet made of, how is it changing and how does it affect your life?

Authors:

Abstract

The scope of this article is to analyse net neutrality. It starts by contextualizing internet’s birth, its purposes, the main principles used on its design and how was the original architecture of that network, specially its modular structure and layers. Next, it describes some recent changes on the internet’s structure, highlighting the monitoring and control of data packets, by accessing its headers or using a technology known as DPI (deep packet inspection), which allows the monitoring of the content of each packet, in order to discriminate between them. This part explains how this subject affects the life of billions of people, worldwide. Finally, the paper focuses on net neutrality, as the main legal response to the previously described problems, explaining how it derivate from the original design principles of internet, in order to support a conclusion that reconciles, at one side, free competition and investment strategies, and, at the other side, privacy, freedom of speech, innovation and internet users decision-making autonomy.
LEGAL CONVERSATIONS
BETWEEN ITALY AND BRAZIL
a cura di
Giuseppe Bellantuono
e
Fabiano Lara
2018
QUADERNI DELLA FACOLTÀ DI GIURISPRUDENZA
35
2018
Al fine di garantire la qualità scientifica della Collana di cui fa parte, il
presente volume è stato valutato e approvato da un Referee interno alla
Facoltà a seguito di una procedura che ha garantito trasparenza di criteri
valutativi, autonomia dei giudizi, anonimato reciproco del Referee nei
confronti di Autori e Curatori.
PROPRIETÀ LETTERARIA RISERVATA
© Copyright 2018
by Università degli Studi di Trento
Via Calepina 14 - 38122 Trento
ISBN 978-88-8443-817-1
ISSN 2284-2810
Libro in Open Access scaricabile gratuitamente dall’archivio IRIS -
Anagrafe della ricerca (https://iris.unitn.it/) con Creative Commons
Attribuzione-Non commerciale-Non opere derivate 3.0 Italia License.
Maggiori informazioni circa la licenza all’URL:
http://creativecommons.org/licenses/by-nc-nd/3.0/it/legalcode
Il presente volume è pubblicato anche in versione cartacea, con il
contributo finanziario della Regione Autonoma Trentino-Alto Adige, per i
tipi di Editoriale Scientifica - Napoli, con ISBN 978-88-9391-461-1.
Novembre 2018
LEGAL CONVERSATIONS
BETWEEN ITALY AND BRAZIL
a cura di
Giuseppe Bellantuono
e
Fabiano Lara
Università degli Studi di Trento 2018
TABLE OF CONTENTS
Page
Durval de Noronha Goyos Junior
Foreword .........................................................................................
.
VII
Giuseppe Bellantuono and Fabiano Lara
Preface .............................................................................................
.
IX
Marcelo Andrade Féres
La legge brasiliana anticorruzione e la responsabilità delle per-
sone giuridiche: problemi e prospettive ..........................................
.
1
Nicoletta Parisi
ANAC and the Italian model for the prevention of corruption:
best practices, challenges and gaps ................................................
.
17
Alexandre Travessoni Gomes Trivisonno
Obedience to authority and the relation between law and moral:
Kant as legal positivist? ..................................................................
.
37
Federico Puppo
The concept of “system” in legal theory: a philosophical insight ..
.
57
Mônica Sette Lopes
Judge training for orality: experience, history and memory of the
unwritten law ...................................................................................
.
71
Marcelo de Oliveira Milagres
E-commerce and anonymous consumers and traders .....................
.
93
TABLE OF CONTENTS
VI
Page
Leonardo Parentoni
Network neutrality: what is Internet made of, how is it changing
and how does it affect your life? ......................................................
103
Fabiano Teodoro Lara
Some mechanism design concepts for legal institutions ..................
.
147
Girolamo Domenico Treccani
Populações tradicionais amazônicas: Afirmação e desconstrução
de direitos ........................................................................................
.
165
Giuseppe Bellantuono
Comparative law for what kind of development? ............................
.
189
NETWORK NEUTRALITY:
WHAT IS INTERNET MADE OF,
HOW IS IT CHANGING
AND HOW DOES IT AFFECT YOUR LIFE?
Leonardo Parentoni
SUMMARY: 1. Introduction and delimitation of the subject. 2. Brief history
of the Internet. 3. What is Internet made of? Network design and its layers.
4. How is Internet changing. 5. How does it affect your life. 6. The main regu-
latory response: network neutrality. 7. Internet design and the limits of Net-
work Neutrality. 8. Conclusion.
1. Introduction and delimitation of the subject
This chapter does not only bring legal considerations. Nor it brings
only considerations related to the Information Technology, as the au-
thor has no specific training in this last area. Rather, it seeks to merge
both, understanding that their joint analysis is essential to understand
the actual functioning of the Internet and, with that, propose legal solu-
tions that are plausible and feasible, not only from a theoretical point of
view, but especially in the factual aspect. After all, the best theory is the
one that works in practice.
Based on this premise, it is intended to address the origins of the In-
ternet and its unique features, responsible for turning it into the success
it is today. Afterwards, it is shown how the Internet has been changing
in recent years, and the consequences of this change in people and en-
trepreneurs’ lives and in the proper functioning of markets.
Then it is discussed the aspect that, in the author’s view, is the main
regulatory response to the new problems: the network neutrality. Net-
work neutrality is conceptualized with foreign legal literature and the
treatment of the subject brought by the Brazilian Civil Rights Frame-
LEONARDO PARENTONI
104
work for the Internet is also analysed. Finally, the author presents his
understanding of how should the network neutrality be applied.
2. Brief history of the Internet
The remote origin of the Internet pushes back to the 50s of the last
century, long before its worldwide expansion. Like any other technolo-
gy, it was the result of a socioeconomic context, which directly influ-
enced the form and the goals for which it was developed. Understand-
ing this point is essential to critically analyse the Internet of today1.
Indeed, the world was living the so called “cold war”, a clear con-
flict between the United States and the Soviet Union. During this peri-
od, the fastest means of communication was by telephone. The tele-
phone network, however, was very vulnerable to military strikes.
Aware of this weakness, the US Department of Defense wanted to de-
velop a safer alternative. That’s when a systems programmer at Rand
Corporation, named Paul Baran, suggested, still in the 50s, a computer
network model similar to the Internet. The Department of Defense,
then, consulted with AT&T, the company that monopolized the tele-
phone communication of the country, to figure out if she would be in-
terested in developing this model. AT&T immediately rejected the idea,
affirming that it was not feasible2.
1 M. CASTELLS, La Galaxia Internet, Translation: Raúl Quintana, Barcelona, 2001,
p. 23. “(…) la producción de una determinada tecnología en un momento histórico
condiciona su contenido y los usos que se hacen de ella en su desarrollo futuro. Internet
no escapa a esta regla. La historia de Internet sirve para comprender su evolución poste-
rior”.
2 A.S. TANENBAUM, D.J. WETHERALL, Computer Networks, 5th Ed., Boston, 2011,
p. 55. “Around 1960, the DoD [Department of Defense] awarded a contract to the
RAND Corporation to find a solution. One of its employees, Paul Baran, came up with
the highly distributed and fault-tolerant design (…) Baran proposed using digital pack-
et-switching technology. Baran wrote several reports for the DoD describing his ideas
in detail. Officials at the Pentagon liked the concept and asked AT&T, then the
U.S. national telephone monopoly, to build a prototype. AT&T dismissed Baran’s ideas
out of hand. The biggest and richest corporation in the world was not about to allow
NETWORK NEUTRALITY
105
Then, in October 1957, the Soviet Union took a worldwide reported
step towards technological progress, launching the Sputnik satellite. In
reaction to this, the then US President, Dwight Eisenhower, developed,
in September 1959, within the Department of Defense, an agency for
advanced research projects (Advanced Research Projects Agency -
ARPANET), dedicated specifically to improve the communication nets.
The objective was to overcome the Soviet Union.
Among other purposes, the ARPANET would develop a safer com-
munication network that could remain available even when one of its
components got damaged (eg. in case of destruction of a military base,
the information stored in it would remain accessible)3.
However, at that time, the ARPANET had a modest infrastructure4.
Therefore, it appealed to the academic world to develop their objec-
tives5. Professors and students then rescued the ideas of Paul Baran,
demonstrating the feasibility of establishing a computer network based
on the standard that had been suggested by him years before6.
some young whippersnapper tell it how to build a telephone system. They said Baran’s
network could not be built and the idea was killed”.
3 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 45-46. “Given the DoD’s [De-
partment of Defense] worry that some of its precious hosts, routers, and internet-work
gateways might get blown to pieces at a moment’s notice by an attack from the Soviet
Union, another major goal was that the network be able to survive loss of subnet hard-
ware, without existing conversations being broken off. In other words, the DoD wanted
connections to remain intact as long as the source and destination machines were func-
tioning, even if some of the machines or transmission lines in between were suddenly
put out of operation. Furthermore, since applications with divergent requirements were
envisioned, ranging from transferring files to real-time speech transmission, a flexible
architecture was needed”.
4 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 56. “ARPA had no scientists or
laboratories; in fact, it had nothing more than an office and a small (by Pentagon stand-
ards) budget. It did its work by issuing grants and contracts to universities and compa-
nies whose ideas looked promising to it”.
5 The first academic institutions to take part at the Net were the University of Cali-
fornia, Los Angeles, and the Stanford Research Institute, in October 1969. Right after,
the Utah University and the California University, in Santa Barbara, also took part at it.
6 Highlighting the importance of academics for the development of Internet:
K. HAFNER, M. LYON, Where Wizards Stay Up Late: The origins of the internet, New
York, 1996.
LEONARDO PARENTONI
106
In the 70s there were already several Local Area Networks (LANs)
at the United States of America, but they were not interconnected. For
example, at universities. Or even the private networks BBS type (Bulle-
tin Board Systems)7. The challenge was to develop a universal commu-
nication standard, capable of interconnecting all networks and giving
birth to a “network of networks”. For this purpose, the ARPANET be-
gan to promote research in the field, which resulted, in 1978, in the cre-
ation of the TCP/IP protocols, which are still the standard of the Inter-
net8. Based on them, the various local networks got integrated and
started to allow an exchange of information and academic experiences
never seen before. A few years later, in the late 80s, the number of in-
stitutions, devices and people connected was already so great that the
ARPANET was replaced by a more robust network called NSFNET
(National Science Foundation Network), reaching across the country.
The Internet was taking off from its military origins9.
At this point, it would not be feasible to keep the Internet under the
sole responsibility of the US Government. Then, it was created a pri-
vate company controlled by the State (ANS - Advanced Networks and
Services), to manage the progressive transfer of the network to com-
mercial exploitation (“privatization” of the Internet). This was consoli-
dated in the 90s, when many private companies were already acting as
access providers10. It was also at this time that the network expanded
worldwide. That is, effective internationalization of the Internet and its
commercial exploitation have started about 40 years after the first re-
searches on the subject.
This brief historic intended to demonstrate that the Internet was not
created as a market-led initiative. On the contrary, AT&T, a company
7 A famous BBS in the 80s was BITNET, which stands for “because it’s there
newtwork”, evidencing that this kind of net was already a reality.
8 M. CASTELLS, op. cit., p. 25. “Para conseguir que las redes de ordenadores pudie-
ran comunicar entre ellas, eran necesarios unos protocolos de comunicación estandari-
zados. (…) En 1978, Cerf, junto con Pastel y Cohen, que trabajaban en la University of
Southem California, dividieron el TCP en dos partes, añadiendo el protocolo interredes
(IP) y creando así el protocolo TCP/IP estándar sobre el que aún opera Internet”.
9 ARPANET was definitively shut-down in February 1990.
10 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 60.
NETWORK NEUTRALITY
107
that monopolized the US telephone communications, showed complete
disinterest in the technology in the 50s. In fact, the worldwide web was
the result of a partnership between the US Government and educational
institutions of the country. Its birth is both military and academic11. As
summed up by Manuel Castells12: “the Internet was born in the unusual
crossroad between big science, military research and the libertarian cul-
ture”.
As it will be analysed below, this historical feature was of major
importance for the success of the worldwide web.
3. What is Internet made of? Network design and its layers
In order to build something, one needs to not only define how the
building will be, but also materialize this idea through a project. The
architecture/design is the preliminary step to the implementation of the
work. With regard to computer networks, this step is called the network
architecture or network design13. It is the network architecture that de-
11 L. PARENTONI, Documento Eletrônico: Aplicação e Interpretação pelo Poder Ju-
diciário, Curitiba, 2007, p. 27-29.
12 M. CASTELLS, op. cit., p. 31. “Internet nació en la insólita encrucijada entre la
gran ciencia, la investigación militar y la cultura libertaria”.
13 B. VAN SCHEWICK, Internet Architecture and Innovation, Massachusetts, 2010,
p. 20-21. “[architecture] denotes the fundamental structures of a complex system as
dened during the early stages of product development. Similar to the way the architec-
ture of a house is different from the house itself, the architecture of a system is not the
nal, working system; rather, it is a description of the system’s basic building blocks.
(…) In short, the architecture describes the components of the system, what they do,
and how they interact”.
See also: L. LESSIG, Code: Version 2.0, New York, 2006, p. 24; e A.S. TANEN-
BAUM, D.J. WETHERALL, op. cit., p. 31. “A set of layers and protocols is called a net-
work architecture. The specification of an architecture must contain enough information
to allow an implementer to write the program or build the hardware for each layer so
that it will correctly obey the appropriate protocol”.
Also: R.S. WHITT, A deference to protocol: Fashioning a three-dimensional public
policy framework for the internet age, in Cardozo Arts & Entertainment Law Journal,
v. 31, n. 03, p. 689-768, Jul. 2013, p. 704. “‘Architecture’ is a high-level descriptor of a
complex system’s organization of basic building blocks, its fundamental structures.
LEONARDO PARENTONI
108
fines how the entire infrastructure14 of a computer network will be, af-
fecting its functioning and even the cost and the possibilities for future
modification.
There are several possibilities for network architecture. It will de-
pend on the direction followed by its creators. This direction, in turn, is
given by the design principles15. In the development of the Internet,
there were applied four fundamental principles: 1) Packet switching;
2) Modularity; 3) Network layers; and 4) End-to-end.
Understanding the original design of the Internet is essential to per-
ceive how it has been changing and what are the consequences of these
changes.
Indeed, the basic feature of the Internet is to divide the data into
smaller fractions, called packets, transmitting them separately. Each
packet receives information about who is the sender and what is the
address (addressing). Hitherto, something similar to traditional mail, on
paper. The big difference is that the various packets can travel simulta-
neously by several routes, even changing the route along the way, to
privilege the one that is faster and more effective (switching). Getting
to the destination, the packets are reassembled to form the original data
and then delivered. This principle enables the data transmission even if
one or a few network components are disconnected, congested or by
any reason unavailable16. This was the solution suggested by Paul
How the Internet runs is completely dependent on the implementing software code, its
fundamental nature created and shaped by engineers”.
14 About the concept of infrastructure see: R.D. OLIVEIRA, L. PARENTONI, Uma Ad-
vertência sobre Interoperabilidade e o Artigo 154, Parágrafo Único, do CPC, in Revi-
sta Magister de Direito Civil e Processual Civil, Ano IV, n.º 19, p. 51-73, jul./ago.
2007.
15 B. VAN SCHEWICK, op. cit., p. 23. “A design principle describes known connec-
tions between architectural choices and the characteristics of the resulting architecture.
(…) a design principle describes how to design an architecture for a system with
specic quality characteristics, and, like different versions of a dish, the resulting archi-
tectures will differ depending on the design principles that were used to create them”.
16 L.G. ROBERTS, The evolution of packet switching. Proceedings of the IEEE, New
York, v. 11, p. 1307-1313, Nov. 1978, p. 1307. “[A packet switched network] divides
the input flow of information into small segments, or packets, of data which move
NETWORK NEUTRALITY
109
Baran in the 50s, and subsequently endorsed by the US Government,
through ARPANET. The following figure compares traditional data
communication, via direct cable connection, with packet switching:
It is noticed that the difference is the possibility that the data from
“A” to “B” is delivered, even if the direct connection between them is
compromised, as there is the option of using the route through “C”. On
the Internet, for each route there are hundreds or even thousands of al-
ternatives. The following figure illustrates how the data is broken dur-
ing transmission and then reassembled at the destination:
The second principle applied at the original Internet design was
modularity. Based on it, the network consists of many parts, independ-
ent of each other, called modules
17
. The intention is to reduce to a min-
through the network in a manner similar to the handling of mail but at immensely high-
er speeds”.
17
B. V
AN
S
CHEWICK
, op. cit., p. 38. “Modularity is a design principle that inten-
tionally makes components highly independent (‘loosely coupled’). Components of
LEONARDO PARENTONI
110
imum the dependence of the network as a whole upon each component.
For this reason, every module has two types of information, the visible
and the “invisible”18. Visible information is one that any component of
the network needs to know in order to connect with a particular module.
It should remain unchanged and available to any interested party during
whole life of the network, in order to not disrupt communication be-
tween modules. For example, the VGA port format, used for monitors
connection to PCs, is a visible data. Knowing this information, any
manufacturer is able to produce a monitor that supports this port.
On the other hand, the necessary information for the internal opera-
tion of each module is called “invisible” information. It is usually
known only by the module manufacturer, consolidating its competitive
advantage over competitors. For example, the resolution and mode of
operation of the monitor. As long as the visible information remains the
same, modules with different internal configurations are compatible.
That is, the same PC could connect to many monitors. Another example
is the USB port. Holding the visible information (format of the port and
what is needed to connect to it), any manufacturer can develop a com-
patible product, such as mice, printers, cameras, mobile phones, tablets,
etc. The fact that each product has a different internal setting does not
affect the connection. The great advantage of modularity is to allow
internal improvements in each component of the network with no need
to change its entire infrastructure19.
In turn, communication between the various modules is established
through its internal specifications (protocols)20 and through connection
modular designs are called modules. When designing a modular architecture, system
architects decompose the system in a way that minimizes dependencies among compo-
nents”.
18 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 29. “The fundamental idea is that
a particular piece of software (or hardware) provides a service to its users but keeps the
details of its internal state and algorithms hidden from them”.
19 It steps aside from the “all or nothing” logic, allowing ponctual modifications, as
well as reducing the adaption cost of each modification.
20 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 29. “Basically, a protocol is an
agreement between the communicating parties on how communication is to proceed”.
NETWORK NEUTRALITY
111
to the layers that are immediately higher and lower (services)21. Thus,
network layers are arranged vertically (stack of layers)22 with relative
independence from one another.
Yochai Benkler believes that any communication system should
present at least 3 layers23. Specifically in what regards the Internet,
there is no consensus about how many layers there are and which
would they be, and there is more than one classification24. In this paper,
the description comprising 6 layers was chosen: 1) Physical; 2) Of
Connection (data link); 3) Of Network; 4) Of Data Transport; 5) Of
Applications; and 6) Of Content.
The initial layer from the bottom up, is the physical. She encom-
passes the devices that transmit data in its raw form (raw data), such as
network cables, satellites and mobile phone towers25. Next, comes the
21 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 40. “Services and protocols are
distinct concepts. (…) A service is a set of primitives (operations) that a layer provides
to the layer above it. The service defines what operations the layer is prepared to per-
form on behalf of its users, but it says nothing at all about how these operations are
implemented. A service relates to an interface between two layers, with the lower layer
being the service provider and the upper layer being the service user.
A protocol, in contrast, is a set of rules governing the format and meaning of the
packets, or messages that are exchanged by the peer entities within a layer”.
22 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 29. “To reduce their design com-
plexity, most networks are organized as a stack of layers or levels, each one built upon
the one below it. The number of layers, the name of each layer, the contents of each
layer, and the function of each layer differ from network to network. The purpose of
each layer is to offer certain services to the higher layers while shielding those layers
from the details of how the offered services are actually implemented”.
23 Y. BENKLER, From consumers to users: Shifting the deeper structures of regula-
tion toward sustainable commons and user access. Federal Communications Law
Journal, in George Washington University Law School, v. 52, n. 03, p. 561-579, 2000.
24 Depending on the classification, certain layers are assembled or disassembled.
There are, for example, those who consider that the physical layer would not be a part
of the network itself, as it does not correspond to one of its modules, merely transmit-
ting the data in raw form (raw data).
25 C.S. YOO, Protocol Layering and Internet Policy, in The University of Pennsyl-
vania Law Review, v. 161, n. 06, p. 1707-1771, 2013, p. 1747. “Any path that can con-
vey a message sequence can constitute a link in this layer, whether physical or not. The
means of encoding information varies widely depending on whether the carrier wave is
composed of visible light passing through a fiber optics network, an electromagnetic
LEONARDO PARENTONI
112
connection layer, responsible for establishing the connection between
the devices that transmit the raw data and the network itself26. The third
layer is the one that assigns identification to the data, through the Inter-
net Protocol (IP)27. In effect, any device connected to the Internet must
have an identification number. It is through this number that the IP pro-
tocol individualizes the device, ensuring that it properly sends and re-
ceives data. In a simplified way, the IP address works for the Internet
such as the home address of the person works for the post office. Each
device has a unique IP, but the same person may have multiple devices
connected to the network simultaneously, each one with a different IP.
For example, mobile phone, PC, tablet etc. What matters is to individu-
alize the devices, not their holder28.
After identifying the data with the source and the destination IPs
(end hosts), the fourth layer is responsible for effectively transmitting
them29. This is done through another protocol called TCP (Transfer
Control Protocol). It is this protocol that fraction the original data into
smaller packets and, in the sequence, transmits them.
wave passing through a copper wire, or an electromagnetic wave passing through the
ether”.
26 C.S. YOO, op. cit., p. 1746. “Data-link layer protocols share with network-layer
protocols the responsibility for guiding traffic through the network; as a result, data-link
layer protocols necessarily run in switches as well as hosts”.
27 C.S. YOO, op. cit., p. 1745. “(…) this layer provides the uniform basis that each
network connected to the Internet uses to transmit data communications across an ever-
changing landscape of technologically heterogeneous systems”.
28 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 43. “The network layer controls
the operation of the subnet. A key design issue is determining how packets are routed
from source to destination”.
The possibility of a dynamic IP does not interfere in this reasoning because in every
connection the device will have a specific IP. There will only be no guarantee that the
following connection uses the same number.
29 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 44. “The basic function of the
transport layer is to accept data from above it, split it up into smaller units if need be,
pass these to the network layer, and ensure that the pieces all arrive correctly at the
other end”.
NETWORK NEUTRALITY
113
The IP and the TCP work together, consolidating the heart of the In-
ternet30: one individualizes the data by inserting the source and the des-
tination addresses; the other splits the data, transmits and reassembles
them at the destination. For that reason, it is common to hear “TCP/IP
protocol” (in the singular), when in reality they are two different31 and
complementary things (a protocol suite).
The fifth layer network is the applications one32. On this layer all
software and typical features of the Internet are run (in programming
language), such as the browsers used to access sites, the social networks
and the download of files. The end result of these applications is then
transmitted to the user, by the sixth and final layer. Most users general-
ly think that the Internet is composed only by this last layer, as it repre-
sents all they effectively see on the screen of their devices. The lower
layers are commonly known and exploited only by the system’s pro-
grammers and the services providers.
This modular architecture, organized in layers, defines the way the
data communication via the Internet is processed. Indeed, there is no
direct transfer from a layer, in the source device, to the equivalent layer
in the target device. In fact, the communication first flows “top-down”,
from the device used to generate the data, through TCP/IP, which will
individualize, route and fraction them into packets, until they reach the
physical layer. From there, they will be transmitted in its raw form until
they reach the place in which the receiver is located. After that, they
start to flow “from the bottom up”, ranging from the physical layer to
the connection one, through TCP/IP, which will be responsible to check
if the destination is correct and, if so, reassemble the packets, reconsti-
tuting the original data, to then send them to the application that will
process them (in programming language), until the final result is dis-
30 L.B. SOLUM, M. CHUNG, The Layers Principle: Internet Architecture and the
Law, in Notre Dame Law Review, v. 79, n. 03, p. 815-948, Jan. 2004, p. 839. “Without
TCP/IP there can be no Internet”.
31 In the origins of the Internet it has been proposed to merge both protocols into a
single one, the Internetwork Transmission Control Protocol - ITCP. However, this idea
was soon discarded: B. VAN SCHEWICK, op. cit., p. 96 a 98.
32 C.S. YOO, op. cit., p. 1742. “This layer encompasses a wide variety of protocols,
each designed to support particular classes of applications”.
LEONARDO PARENTONI
114
played on the receiver device
33
. The following figure facilitates the vis-
ualization of this route:
The characteristic feature of this network architecture is that the
core layers, such as TCP/IP, are not able to identify the content of the
transmitted data. For them, it is not a concern whether the data repre-
sent a legal text, a journalistic article or any other manifestation of
thought. The type of application to which they relate is also indifferent.
For example, if they relate to a web page, to the transmission of a mov-
ie or to a post on a social network. The same and universal standard is
followed for the transporting of any data, directed to any kind of appli-
cation
34
. The applications, located at the penultimate upper layer, are
the sole responsible for identifying the content of the data and for what
purpose they should be applied to. This is the fourth and final design
principle used in the construction of the Internet: Running Applications
in the Upper Layer (end-to-end). Opposite option is that one in which
the functioning of the applications is controlled by the central layers of
33
L.B. S
OLUM
, M. C
HUNG
,
op. cit., p. 816-817.
34
B. V
AN
S
CHEWICK
, op. cit., p. 72. “[This] results in a network that is not able to
distinguish between the different applications running over it, or to control or to posi-
tively or negatively affect their execution”.
NETWORK NEUTRALITY
115
the network (core centered). This option was not used on Internet’s
original design.
The simplicity in the operation of the central layers of the Internet,
coupled with the absence of filters, allows any new application to be
immediately and worldwide disseminated, without requiring any further
adjustment in the basic network infrastructure35. For this reason, the
Internet has become a major conductor, open to the transmission, shar-
ing and collective construction of various types of content, stimulating
innovation and dialogue36. The success of the Internet, therefore, is
closely related to the original design of the network.
On the other hand, as the internal specification of each module re-
mains visible only to its manufacturer (hence hidden from all other ser-
vices that connect to this module), the integration between them is not
the best possible. There are, thus, loss of performance in the modular
architecture model, if compared with the full integration of the network
components.
These advantages and disadvantages of the original Internet archi-
tecture will be discussed with further detail later. For now, it is enough
to highlight that there is no perfect alternative, capable of providing
only qualities. The network design necessarily involves a trade-off be-
tween losses and gains.
35 B. VAN SCHEWICK, op. cit., p. 140. “In an end-to-end network (such as the origi-
nal Internet) in which network operators and application designers follow the broad
version of the end-to-end arguments, designing, implementing, testing, and deploying a
new application do not require changing the network’s core (…)”.
Obviously, this rule is not absolute. Certain applications, to present an acceptable
performance, may eventually require structural changes in the network. This is the case
of streaming of audio and video, as well as the peer-to-peer applications.
36 S. RANCHORDÁS, Does Sharing Mean Caring? Innovation in the Sharing Econo-
my, in Minnesota Journal of Law, Science & Technology, v. 16, n. 01, p. 01-63, winter
2015, p. 14-15. “Innovation is a broad concept that can be defined differently depend-
ing on the context and field in question. (…) innovation is defined as ‘the ability to take
new ideas and translate them into commercial [or effective social] outcomes by using
new processes, products, or services (…)’. Innovation is more than an idea or a novelty;
it must be the first successful concretization of an idea in the marketplace or in society”.
LEONARDO PARENTONI
116
4. How is Internet changing
In the previous topics, it has been demonstrated that the Internet is a
human construction, guided by certain principles and objectives37. Like
any other human construction, it can be altered. After all, there are oth-
er possible settings, each one guided by different social, political and
economic purposes38. It happens that, depending on the type of change,
and on the intensity with which it is implemented, this can give rise to a
new Internet, entirely different from the one the world has used to
know. This topic intends to, briefly, point out some of these changes. In
the next topic it will be done a critical analysis about them, in the legal
perspective.
It is known that the phone lines were built focusing, primarily, at the
transmission of sounds. Just as TV cables were designed only to the
service of paid TV. None of these technologies were designed to the
Internet typical data flow, especially because the Internet was devel-
oped later. Consequently, the transmission speed for the Internet appli-
cations was originally very low. Subsequent changes in these instru-
ments, however, have made possible the transmission of data at speeds
until then unforeseeable, which rendered these new technologies the
name “broadband”39. Thus, copper cables were gradually replaced by
optical fibre, whilst traditional phone lines evolved to ADSL (asymmet-
ric digital subscriber line)40. The same has occurred in relation to the
37 L. LESSIG, Code: Version 2.0, cit., p. 06. “We can build, or architect, or code cy-
berspace to protect values that we believe are fundamental. Or we can build, or archi-
tect, or code cyberspace to allow those values to disappear. There is no middle ground.
There is no choice that does not include some kind of building. Code is never found; it
is only ever made, and only ever made by us”.
38 R. LEMOS, Uma Breve História da Criação do Marco Civil, in N. DE LUCCA,
A. SIMÃO FILHO, C.R.P. DE LIMA (Coord.), Direito & Internet III: Marco Civil da In-
ternet - Lei nº 12.965/2014, São Paulo, 2015, t. I, p. 79.
39 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 95-111.
40 In Brazil, for instance, the companies “Oi” and “GVT” provide access via ADSL,
whilst “NET” and “SKY” use the optical fibre cable system. In less urbanized regions
of the country, where telephone networks and cabling are still precarious, alternative
access technologies, through radio or satellite, are often used.
NETWORK NEUTRALITY
117
Internet access through the electricity transmission networks, satellites,
radio waves, microwaves or WiFi (wireless fidelity).
Each of these technologies, in their own way, and for different
backgrounds, allowed the improvement of the Internet-trough commu-
nications. The common element that have to be highlighted is that all of
them could be implemented only in the physical and in the connection
layers, without the need of entirely altering the network operation.
Hence, Internet-connected devices remained fully interoperable. This
was only possible thanks to the original layered design and to the mod-
ularity.
Throughout the evolution of the Internet, several other modifications
were made solely in the application layer, also preserving the core of
the network. There were created, for instance, protocols for the direct
file transfer between devices (file transfer protocol - FTP) and the
communication via e-mail (simple mail transfer protocol - SMTP).
One of the most impactful change in how the Internet is used was
the development of the DNS (domain name system)41. It has been al-
ready said that every device connected to the network has an IP ad-
dress. To access the device, it would then be necessary to type this ad-
dress, consisting in numerical sequences, such as “200.251.4.1”. How-
ever, such sequences are difficult to memorize and they may change
over time. This made surfing at the Internet extremely complex. It was
difficult for people to memorize a sequence like this for each site they
commonly accessed, and check if the numbers had not changed.
In order to solve this problem, there was an application capable of as-
sociating each IP address to a specific name, so that typing the name re-
placed typing the numeric address. This application is called DNS. She is
responsible for the fluid and intuitive way navigating on the network cur-
rently is, by simply typing “www.google.com” or “www.facebook.com”
to access sites, whatever their IP numbers are. The DNS was not part of
the original architecture of the network and was introduced later on,
shortly before its commercial exploitation.
41 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 611-612.
Also: R.F. MACIEL, A Requisição Judicial de Registro de Conexão e Aplicações no
Marco Civil, in N. DE LUCCA, A. SIMÃO FILHO, C.R.P. DE LIMA (Coord.), Direito &
Internet III: Marco Civil da Internet - Lei nº 12.965/2014, São Paulo, 2015, t. II, p. 478.
LEONARDO PARENTONI
118
Another important recent change, which also has been unnoticed by
most Internet users, is the transition from IPv4 to IPv6. Indeed, IPv re-
fers to the IP protocol version that, as described, serves to individualize
devices on the network by assigning them an IP address. When the In-
ternet was created, the version number 4 of this protocol (IPv4) was
adopted42 and, in accordance to it, addresses were formed by four nu-
merical sequences, each one ranging from 0 to 255. For example,
“200.251.0.1”. This version allowed approximately 4.29 billion devices
to be simultaneously connected. It may seem a lot, but it is not. Know-
ing that the tendency is that more and more individuals and companies
have access to the Internet, and that each one can consume alone, tens
or even hundreds of IP addresses, then it follows that the total number
of addresses was doomed to someday end43.
When this comes to happen, the Internet would be “crowded”, pre-
venting the addition of new users and/or devices44. To overcome this
problem, another version of the IP protocol was developed, called
IPv645. Through IPv6 it is possible to offer, simultaneously, a few tril-
lion times the maximum number of addresses supported by IPv4. An
impressive amount, which will be hardly overwhelmed, even if the
number of network-connected devices increases46.
42 Technical description at UNITED STATES OF AMERICA, The Internet Engineering
Task Force - IETF, available at <https://tools.ietf.org/html/rfc791>, access 21 March
2016.
The three previous versions were used in the network early days, when it was still
restricted to the US.
43 This could take decades longer to happen if certain IP addresses had not been as-
signed, collectively, to some US institutions, far above their needs.
44 Phenomenon dubbed IPcalypse in a jocular allusion to the Apocalypse of the In-
ternet, available at <http://www.ipcalypse.net/>, access 21 March 2016.
45 Technical description at UNITED STATES OF AMERICA, The Internet Engineering
Task Force - IETF, available at <https://tools.ietf.org/html/rfc2460>, access 21 March
2016.
46 It is estimated that IPv6 addresses allows to form more addresses than the total
number of atoms in the universe: A. MELANCIA, O crescimento da Internet, in Revista
Programar, available at <http://www.revista-programar.info/artigos/ipv6-para-progra
madores/>, access 09 April 2016.
NETWORK NEUTRALITY
119
IPv6 is based on the hexadecimal system, combining letters and
numbers. This new protocol is being implemented gradually, coexisting
with IPv4. For example, the email address “www.facebook.com” corre-
sponded, in March 2016, to the IPv4 “200.175.89.139” and to the IPv6
“2A03: 2880: 2130: CF24: face: b00c: 0: 25de66.220.158.68”47.
Hitherto it has been pointed out some changes that are happening at
peripheral layers of the Internet (physical, connection and applications
layers). None of them, however, reached the heart of the network, to the
point of changing the principles of the original design, such as the oper-
ation of TCP/IP protocols. It happens that this kind of change – much
more serious and controversial – is also already happening. It consists
on the introduction of technologies for monitoring the package content
itself (deep packet inspection - DPI)48.
The DPI is a software capable of examining both the packet rout-
ing’s information (headers), which indicate its source and destination,
and the actual content of each data packet. This occurs when the data
are passing through a given point in the network, where the software is
programmed to act49. From this point on, it is possible to discriminate
packets, defining whether they should proceed or be discarded, as well
as imposing different speeds for each transmission. The access to head-
ers is necessary and legal, because it is essential to the flow of data on
Internet. What the law forbids is the discrimination between packets,
based on this information.
This was inconceivable in the origins of the Internet. As seen, the
worldwide web was designed under the principle of modularity, to en-
sure interoperability with any new program or hardware, without re-
quiring changes to the network infrastructure (the original modules) or
47 The use of dynamic IPs was not taken into account, to simplify the example.
48 M. GEIST, The Emergence of Net Neutrality Regulation in Canada: How Canada
Developed a Consensus Policy on One of the Internet’s Most Contentious Issues, in
N. DE LUCCA, A. SIMÃO FILHO, C.R.P. DE LIMA (Coord.), Direito & Internet III: Marco
Civil da Internet - Lei nº 12.965/2014, São Paulo, 2015, t. II, p. 646. “The DPI [deep
packet inspection] capabilities allow ISPs to identify the type of content that runs on
their networks and render it possible for them to manage the traffic based on the con-
tent”.
49 Such programs are normally operated by the Internet service providers.
LEONARDO PARENTONI
120
to its layers. As long as the external information of each module re-
mained the same, an infinite range of new modules could be connected.
Not only that, but the TCP/IP protocols worked neutrally, transmit-
ting any data packets, regardless of their origin, destination or of which
given species they were. Only at the top layer of the network that data
would be identified, by each particular application, in order to generate
the final result (a movie, music, text etc.). During the transmission, it
shall be reinforced, it was absolutely irrelevant to which sort of data it
was.
It turns out that this has changed. Currently, much of the physical in-
frastructure of the Internet belongs to the private sector50. Thus, it is no
surprise that private interests are behind the major recent changes in the
network operation. One of the most controversial is precisely the intro-
duction of the DPI. From this technology on, the Internet service pro-
viders have gained a previously inconceivable power. They are able to
manipulate the way users sense the network, providing greater speed to
some data packets over others. Or even interrupting the transmission of
certain packets. Taking up the example of the Post Office, it is as if
providers could read all the posted correspondence, choosing whether
and when to deliver each one. The recipient would tend to think that
some services were more efficient than others, as some of the corre-
spondence relating to them (for example, payment slips), always arrive
first, whilst others tend to delay. Now picture the range of business,
political and ideological interests that could abide such discrimination.
Practices of this sort, by the Post Office, would sound absurd and
unacceptable. However, on the Internet, they are already taking place,
and with great frequency. Worst of all, many users do not even know
50 L.F.T. VERGUEIRO, Marco Civil da Internet e Guerra Cibernética: Análise Com-
parativa à Luz do Manual de Talin Sobre os Princípios do Direito Internacional
Aplicáveis à Guerra Cibernética, in N. DE LUCCA, A. SIMÃO FILHO, C.R.P. DE LIMA
(Coord.), Direito & Internet III: Marco Civil da Internet - Lei nº 12.965/2014, São
Paulo, 2015, t. II, p. 634-635. “(…) The core of the Internet is formed by a series of
infrastructure elements: the interconnection points (IXP), the backbones or national
stems, regional networks and local networks, in general, are in owned by private enti-
ties, in a way that its openness and its technology contains elements with potential for
internal control”.
NETWORK NEUTRALITY
121
it51. This structural change in Internet has misrepresented its main fea-
ture, which was precisely the neutrality of the TCP/IP protocols in what
regards the transmission of content. In the original design of the net-
work, the packet identification was restricted to the upper layer applica-
tions, where each data packet was assigned for a specific purpose (end-
to-end principle). The core of the network, composed by the TCP/IP
protocols, worked neutrally, without worrying about the content of the
packets or who would be their sender and recipient.
The next topic will seek to demonstrate how this structural change
in the core network layers leads to very serious and worldwide conse-
quences, both micro and macroeconomic.
5. How does it affect your life
But after all, how this issue affects your life?
The subject of this text not only interest large corporations, even
though it also interests them. Neither it affects solely the systems pro-
grammers or the engaged Internet users (known as “activists”). It af-
fects the lives of all people who use, have used, or someday might
make usage of the Internet52. In essence, billions of people around the
world. This subject affects, indirectly, even those who do not have ac-
cess to the World Wide Web53. The problem has micro and macroeco-
nomic outcomes, both in country’s internal and international levels,
51 M. GEIST, op. cit., p. 650. “Yet, the public generally remained somewhat apathet-
ic towards the issue, with the net neutrality debate crowded by ‘telecom lobbies’”.
52 Only in Brazil, about 86 million people frequently use the Internet: V. CAPUTO,
Mais da metade dos brasileiros são usuários da internet. Caderno Tecnologia, in Revi-
sta Exame, available at <http://exame.abril.com.br/tecnologia/noticias/mais-da-metade-
dos-brasileiros-sao-usuarios-da-internet>, access 07 April 2016.
53 Imagine a resident of an underdeveloped country, who has never accessed the In-
ternet, but that could be hired as an employee of a startup, whose business model is
based on the Internet. The income from this job would have been a means of improving
the living conditions of this citizen and his family. However, he may never have this
chance, once the structural changes of the Internet can make the aforementioned startup
not exist.
LEONARDO PARENTONI
122
being able to meddle with the organization and functioning of mar-
kets54.
In fact, the network architecture has direct influence in the costs of
possible alterations of it in the future55. In the case of Internet, its origi-
nal design used to provide low costs for innovations, because new
products or services, as a rule, would require only changes in the top
layer, on the applications, remaining intact all others56. Of course this
rule is not absolute, since certain types of innovation in the application
layer also require changes in the lower layers of the network, to ensure
good quality of the services. For example, the transmission of video in
real time (streaming), has required certain improvements in the physi-
cal and connection layers, so that the increase in the data stream would
not overload the network. Even in the TCP/IP layers it would require,
eventually, changes in the protocols, in order to adapt them to new ap-
plications. IPv6, mentioned above, is a good example. What should be
clear is that, in all these cases, the original network design – and the
principles that guided its creation – has been preserved.
By contrast, in networks that have integrated design, the cost of any
alteration tends to be higher, since it is not enough to change just one or
54 F.K. COMPARATO, Capitalismo e Poder Econômico, in Revista da Faculdade de
Direito da UFMG, special number: em memória do Professor Washington Peluso Albi-
no de Souza, p. 167-195, 2013, p. 169. “What is properly called market, and that have
always been the center of the attention of classical political economy, is the open space
of distribution of goods and the provision of services, where imperates the law of divi-
sion and specialization of tasks”.
55 B. VAN SCHEWICK, op. cit., p. 116. “(…) the number and the kinds of compo-
nents that need to be changed to realize a particular innovation may differ among archi-
tectures. As a result, realizing that innovation may be more expensive in one architec-
ture than in another”.
56 B. VAN SCHEWICK, op. cit., p. 151. “The Internet was designed to interconnect
subnets regardless of their physical network technology. (…) To connect a new type of
subnet to the Internet, all one needs is a router that can connect this type of subnet to a
physical network that uses an existing network technology. (…) These design choices
make it possible to innovate on application-layer protocols (as long as no other applica-
tion-layer protocols depend on them) and on link-layer protocols (as long as the IP ser-
vice interface stays the same) without any system-adaptation costs, both when new
protocols are added and when existing protocols are innovated upon”.
NETWORK NEUTRALITY
123
a few modules, being necessary to adapt the entire infrastructure57. And
the higher the cost, the more difficult it becomes to small entrepreneurs
to launch new products or services. Consequently, integrated networks
present two disadvantages: 1) they make innovation more expensive;
and 2) they tend to concentrated the innovation in a small group of peo-
ple, precisely the ones that have sufficient resources to bear the costs of
major changes in the network infrastructure.
Thus, the architecture of a computer network certainly influences
who is able to innovate. But it doesn’t stop there. It also affects some
other aspects. Considering that the costs of innovation are prohibitive
for small entrepreneurs, it tends to concentrate the innovation only on
large companies such as Google, Facebook, Apple and Microsoft. So
these companies’ interests are the ones that will guide what should be
created and when58.
Considering only fully integrated network architecture, creations
that today bring immense comfort and satisfaction to users, such as
WhatsApp and Skype, would probably not exist. What these applica-
tions have in common is the fact that they have emerged as an alterna-
tive to traditional communication, just as telephone, which is why they
have faced strong resistance from mobile telephony operators. As it can
be assumed, those operators would not have developed something that,
at the time, would conflict with their own business model.
Thus, the creation of these applications was only possible thanks to
the original design of the Internet. In this model, the creative power is
shifted from the center (large companies) to the end (the final users). As
long as any user can make use of the pre-built network infrastructure,
57 B. VAN SCHEWICK, op. cit., p. 121. “Owing to the different levels of coupling be-
tween components, the costs of realizing a specic innovation probably will be greater
in an integrated architecture than in a modular architecture”.
58 P.H.S. RAMOS, O Marco Civil e a Importância da Neutralidade de Rede: Evidên-
cias Empíricas no Brasil, in N. DE LUCCA, A. SIMÃO FILHO, C.R.P. DE LIMA (Coord.),
Direito & Internet III: Marco Civil da Internet - Lei nº 12.965/2014, São Paulo, 2015,
t. II, p. 149. “In an extreme model, the strengthening of the core of the architecture can
lead to a similar communication model, of what happened in traditional television –
even if users have the choice to change the channel, the communication flow has pri-
marily one direction, and the decisions on content availability and the use of applica-
tions will be restricted to the interests of those who manage the core network”.
LEONARDO PARENTONI
124
without the need for adjustments, it becomes considerably easier and
cheaper to develop new products or services and, immediately, to make
them available to the market, possibly even worldwide. Mark Zucker-
berg, founder of Facebook, for example, payed the first servers that
hosted the famous social media for only $85.00 per month59.
Even when external financing is needed, the modular architecture
seems to be more advantageous. It allows the capture of funds to be
done in an alternative form, other than the financial market, and also
the reduction of costs, such as in crowdfunding60. This gives greater
freedom and flexibility to developers, since they don’t need to submit to
external interferences in their ideas and business plans, what would in-
evitably occur if they were to depend on investments coming from in-
vestment funds or large companies.
The original design of the Internet, that provides the capability of
innovation described above, even against the will of the large compa-
nies in the sector, is suffering several changes. And the current changes
are much deeper than those experienced in the past. Depending on the
type and the way they are consolidated, they may compromise: 1) in-
novation; 2) freedom of expression; 3) privacy; and 4) the proper func-
tioning of markets.
In this context, if the Internet had not been designed based on packet
switching, modularity, layering and running applications in the upper
layer, it certainly would not have become what it is today. Features and
applications that the world is now used to, simply would not exist. And
worst of all, people would not notice or miss them, because they
wouldn’t have even tried these features…
59 B. VAN SCHEWICK, op. cit., p. 206.
60 N.M. MARTINS, P.M.B.P. DA SILVA, Funcionalidade dos sistemas financeiros e o
financiamento a pequenas e médias empresas: o caso do crowdfunding, in Revista
Economia Ensaios, v. 29, especial edition (Associação Keynesiana Brasileira), p. 25-
56, Dez. 2014, p. 26. “(…) It can be said that crowdfunding is an alternative form of
financing, that connects, directly through the Internet and social networks, those that
can offer, lend or invest resources with those in need of funding for projects or specific
business. Moreover, the collective funding is characterized by the fact that the projects
and businesses mentioned are financed by small contributions of a large number of
individuals, anonymously”.
NETWORK NEUTRALITY
125
An example illustrates this perfectly. Today, the peer-to-peer
(P2P)61 technology is consolidated and provides easy file sharing, for
many different purposes62. However, when it was developed, it suffered
harsh attacks and it only survived thanks to the original design of the
Internet.
The first online platform that allowed users to share free music,
worldwide, was Napster63. This application was strongly based on P2P
technology. With its success and rapid growth, Napster was sued by the
US recording industry, since the songs were being transferred without
paying copyrights64. This lawsuit resulted in the extinction of the com-
pany. Despite of that, the P2P technology was not affected. The case
preserved the original configuration of the Internet, since its effects
were restricted to the upper layer of the network, specifically affecting
only the Napster application.
Because of that fact, the P2P technology remained available and,
through it, were developed several other products and services with
huge success and without the Napster illegality vices. Skype is one of
them. Other services highly based on technology still bring legal chal-
lenges such as Uber65 and Airbnb66. But it is a fact that it would not
61 R. KRISHNAN, M.D. SMITH, R. TELANG, The Economics of Peer-To-Peer Net-
works, in Journal of Information Technology Theory and Application, v. 05, n. 03,
p. 01-24, 2003, p. 01. “P2P networks allow a distributed community of users to share
resources in the form of information, digital content, storage space, or processing ca-
pacity. The novel aspect of these networks is that, in contrast to client-server networks
where all network content is located in a central location, P2P resources are located in
and provided by computers at the edge of the network (…)”.
62 Some of these purposes, of course, are illegal. However, this does not mean that
P2P technology is harmful. After all, every technology tends to be neutral. The usage
that we make of them is what makes it good or bad.
63 Further details can be found at T. WU, When code isn’t Law, in Virginia Law Re-
view, v. 89, n. 04, p. 679-751, Jun. 2003.
64 A&M Records, Inc. v. Napster, Inc. Case n° 00-16401. 239 F.3d 1004. United
States Court of Appeals for the Ninth Circuit. Decided in October, 2, 2000.
To learn more about this case: L. PARENTONI, Documento Eletrônico: Aplicação e
Interpretação pelo Poder Judiciário, cit., p. 178-184.
65 Open content. Wikipedia: the free encyclopedia. “Uber”, available at <https://pt.
wikipedia.org/wiki/Uber_%28empresa%29>, access 03 April 2016. “Uber is an Ameri-
can multinational corporation of urban private transport, based on disruptive technology
LEONARDO PARENTONI
126
have been possible the emergence of any of them if, because of the
Napster case, the infrastructure of the Internet, in its lower layers, had
been affected in order to stop the P2P.
However, changes like that are currently taking place with enough
intensity and speed. The belief that the unique features of Internet will
always remain the same has already been surpassed67. It is undeniable
that, in recent years, it has undergone changes with the clear purpose of
enabling greater control and monitoring68. One of the most aggressive
changes is the interference in the TCP/IP protocols, replacing the origi-
nal feature of neutrality for monitoring data packets, not only in its
origin and destination, but also in relation to the actual content of each
package. A side effect of this is the possibility of differentiating the data
stream, according to certain interests.
In the case of China, for example, those interests are political in na-
ture and seek to institutionalize censorship. This way, the Chinese gov-
ernment interferes in the network layers in order to block access to any
sites that, supposedly, contain “inappropriate content”, which means
networked through an e- hailing app that offers a similar service to the traditional taxi,
popularly known as ‘paid ride’ services.
(…) E-hailing is the act of ordering a taxi through an electronic device, usually a
mobile phone or smartphone. It replaces traditional methods to call taxis, such as phone
or simply wait or go looking for a taxi on the street”.
66 Open content. Wikipedia: the free encyclopedia. “Airbnb”, available at
<https://pt.wikipedia.org/wiki/Airbnb>, access 03 April 2016. “Airbnb allows individu-
als to rent the entire place or part of his own home, as a form of extra accommodation.
The site provides a search platform and reserves to occur between the person who pro-
vides the accommodation and the tourist who search for rental. It covers more than
500,000 adds, in over 35,000 cities and 192 countries”.
67 About this topic: J.P. BARLOW, A Declaration of the Independence of Cyber-
space, available at <https://projects.eff.org/~barlow/Declaration-Final.html>, access 28
December 2015.
68 As an example: L. LESSIG, The Law of the Horse: What cyberlaw might teach, in
Harvard Law Review, n. 113, p. 501-549, Dec. 1999; ID., Code: Version 2.0, cit.;
J. ZITTRAIN, The Future of the Internet: And how to stop it, London, 2008; S. RODOTÀ,
A Vida Na Sociedade da Vigilância, Tradução: Maria Celina Bodin de Moraes, Rio de
Janeiro, 2008.
NETWORK NEUTRALITY
127
any political content contrary to the government’s own interests69. This
is possible since all the big connection providers of international net-
works (backbones)70 in China, are all subject to strict governmental su-
pervision and control. These providers ensure that IP addresses from
China remain blocked from accessing certain foreign sites, especially
newspapers and news websites71. In a jocular analogy with the famous
Walls of China, this practice became known worldwide as The Great
Firewall of China72.
Not only the states have interfered in the functioning of the Internet.
The market also does that. Major international players such as Google,
Microsoft, Facebook, Apple and other of the same gender continuously
monitor the network usage. The intention is to identify consumption
habits and, with that information, to draw a profile of Internet users that
is able to facilitate the sale of products and services completely focused
on them73. In this context, the data acquired considerable economic val-
ue74, being primarily responsible for paying services that are apparently
69 R. LEMOS, op. cit., p. 79. “The way the network is legally treated in different
countries has very different orientations. And not always democratic values guide that
relationship. In this context, the Internet is not a network disconnected from the existing
political systems. Instead, more and more politics and law have influence in the evolu-
tion of the Internet and how it is governed and administered”.
70 A.S. TANENBAUM, D.J. WETHERALL, op. cit., p. 64. “At the top of the food chain
are a small handful of companies, like AT&T and Sprint, that operate large internation-
al backbone networks with thousands of routers connected by high-bandwidth fiber
optic links. These ISPs do not pay for transit. They are usually called tier 1 ISPs and are
said to form the backbone of the Internet, since everyone else must connect to them to
be able to reach the entire Internet”.
71 To know if a site is blocked in China, it is possible to test it, typing: Great Fire-
wall of China, available at <http://www.greatfirewallofchina.org/>, access 04 April
2016.
72 L.B. SOLUM, M. CHUNG, op. cit., p. 896-910.
73 About this topic: S. BAKER, Numerati, Tradução: Ivo Korytowski, São Paulo,
2009.
74 A. DE FRANCESCHI, M. LEHMANN, Data as Tradeable Commodity and New
Measures for their Protection, in The Italian Law Journal, v. 01, n. 01, p. 51-72,
Mar. 2015, p. 51. “Information, particularly important, significant and relevant infor-
mation, (…) is today’s ‘digital gold’”.
LEONARDO PARENTONI
128
free, such as e-mail accounts or social networks. In all these cases, the
payment is made indirectly. Instead of paying cash, the traditional way,
it allows service provider to have access to user privacy and to make
profit from it75. It is the economic interest guiding the discrimination of
the data stream.
Another form of discrimination based on economic interest is traffic
shaping. It consists in increasing the speed and quality of the service
and, at the same time, reducing the competitors, to damage the latter.
This causes the end user the false impression that a service is better
than other. Imagine, for example, an Internet connection provider (re-
sponsible for the speed and connection quality) which is also – directly
or through a company of the same economic group – the provider of
streaming video service. Thanks to the changes that the Internet has
had, in recent years, this provider can reduce the access speed of its
members for competing services, affecting their quality in order to in-
duce them to use the service of the access provider itself. In Brazil,
NET (connection provider) could increase the speed of the Now service
(its video streaming) and, at the same time, to reduce the speed and
quality of Netflix (main competitor). In this context, subscribers of
The “Letter of Fundamental Rights of the European Union” has been amended to
emphasize the importance of personal data in this new context: EUROPEAN UNION, Eu-
ropean Parliament and Council of the European Union. Letter of Fundamental Rights
of the European Union, Strasbourg 12 December 2007, available at <http://eur-
lex.europa.eu/legal-content/PT/TXT/?uri=celex:12012P>, access 04 April 2016.
75 M. LEONARDI, Marco Civil da Internet e Proteção de Dados Pessoais, in N. DE
LUCCA, A. SIMÃO FILHO, C.R.P. DE LIMA (Coord.), Direito & Internet III: Marco Civil
da Internet - Lei nº 12.965/2014, São Paulo, 2015, t. I, p. 528. “(…) As is known, the
advertisement is directed, which was made possible by the treatment of user data –
personal or otherwise – that sustains the system of services and the free information
online. Other business models subscriptions, micropayments, closed sites – did not
have the same success with the majority of users, that were used to ‘all free’ online.
Between paying a small fee for access, per day or per month, or giving personal data,
almost all users prefer to pay with their data.
Paying with data is a valid choice and must be respected, is a model that allows all
users to participate in the online system, not just those who have the resources to pay
for content and services”.
Having in mind that reality, today it is said that we lived in the Big Data era:
V. MAYER-SCHÖNBERGER, K. CUKIER, Big Data, New York, 2013.
NETWORK NEUTRALITY
129
NET would have the false impression that the Now is much better than
Netflix. In other countries, such practices have resulted in bitter judicial
disputes76.
Another kind of traffic shaping occurs when the Internet access pro-
vider deliberately reduces the speed of certain applications, not to bene-
fit with service that it also provides, but simply because they want to
difficult the use of these applications. This is very common in relation
to P2P. The difference from the Napster case is that now it is not only
about simply combating some applications based on this technology,
using the argument that it stimulates fraud. The proper functioning of
the Internet has been changed to harm the whole P2P, whatever are the
applications that are using it or their purposes.
To justify this, the argument of the access providers is that P2P is
typically used to transmit large data packets, which could overload the
network, decreasing the speed available for other applications. As met-
aphor, the Internet could be seen as a great avenue. The more cars trav-
eling at the same time, the slower the general flow of traffic would tend
to be. The cars in this metaphor would be the data packets and P2P
would represent the slow trucks that catch the flow of other vehicles.
The following topics will show that this argument is false77.
For now, it is important to note that the traffic shaping clearly sub-
verts the original features of the Internet. It tends to accentuate the for-
mation of monopolies and the concentration of economic power, reduc-
ing the innovations and the sharing of content, especially those coming
76 In the US, for example, the country’s largest connection provider (Comcast) arti-
ficially reduced the speed of the main competing service (Netflix), in order to induce
customers to opt for the service provider itself (NBC).
The case took to a lawsuit that ended with an agreement in which Netflix had to pay
extra values to Comcast, so the company would not reduce the speed of access of its
members to the service provided by Netflix. Following, the other competitors were
forced to seek similar agreements with Comcast.
USA. Comcast Corp. versus F.C.C. United States Court of Appeals for the District
of Columbia. 600 F.3d 642, j. 06.04.2010.
77 To avoid the overload of the network, there are alternatives that do not lead nec-
essarily to the P2P restriction. The usage of symmetrical bandwidth, replacing the cur-
rent asymmetric band model, would be one of them: B. VAN SCHEWICK, op. cit., 2010,
p. 69-70.
LEONARDO PARENTONI
130
from small developers. It also means indirect censorship, once the user
finds himself prevented from using certain applications, with quality
and speed that could be expected of them, because of unilateral choice
of the connection providers. This has a direct impact on who will be
able to innovate, what will be created, when there will be innovation
and based on what values/goals. It affects, therefore, the lives of bil-
lions of people and the functioning of the Internet, worldwide. How is
the law reacting to all of this? This is the subject of the next topic.
6. The main regulatory response: network neutrality
The change in the Internet infrastructure is attracting attention from
various fields of science, including the legal area, considering the
amount of consequences it can cause. The main legal response to this
problem is known as the principle of net neutrality78. According to this
principle, the Internet service providers – especially access providers
(ISPs)79 – can not, within certain limits, discriminate the data packets
that travel through its infrastructure, because of the origin, destination
or content of the data. Thereof, this measure seeks to preserve the full
78 T. WU, Network Neutrality, Broadband Discrimination, in Journal of Telecom-
munications and High Technology Law, v. 02, n. 01, p. 141-176, Fall. 2003, p. 165.
“[network neutrality exists to] to forbid broadband operators, absent a showing of harm,
from restricting what users do with their Internet connection, while giving the operator
general freedom to manage bandwidth consumption and other matters of local concern.
The principle achieves this by adopting the basic principle that broadband operators
should have full freedom to ‘police what they own’ (the local network) while re-
strictions based on inter-network indicia should be viewed with suspicion”.
M. GEIST, op. cit., p. 641. “While the definition of net neutrality is open to some
debate, at its core is the commitment to ensuring that Internet service providers (ISPs)
treat all content and applications equally with no privileges, degrading of service or
prioritization based on the content’s source, ownership or destination. (…) Adopting a
neutral approach, in other words, requires strict adherence to one cardinal rule: that
ISPs transport data without discrimination, preference, or regard for content”.
79 M. LEONARDI, Responsabilidade Civil dos Provedores de Serviços de Internet,
São Paulo, 2005, p. 23. “The ISP is the company that provides, as a service, the access
to the Internet for their customers. Typically, these companies have a connection to a
backbone or operate their own infrastructure for direct connection”.
NETWORK NEUTRALITY
131
access to information and the decision-making autonomy of Internet
users, in a way that they are not illegally manipulated to opt for a par-
ticular product or service.
The net neutrality is a worldwide controversial subject80. In Brazil,
the embryo of this principle was already included in the General Tele-
communications Law81. After long and heated debates, the issue was
also included in the Brazilian Civil Rights Framework for the Internet,
in April 2014. This law included the net neutrality as a principle on the
use of the Internet in Brazil, dedicating a specific section to this topic:
Article 3. The discipline of Internet use in Brazil has the following
principles:
(…)
IV - preserving and safeguarding network neutrality;
Article 7. Access to Internet is essential to the exercise of citizenship
and users are assured the following rights:
(…)
V - to the maintenance of the hired quality of Internet connection;
VI - to clear and complete information contained in the services con-
tracts, with details on the arrangements for protecting the connec-
tion logs and access records to Internet applications, as well as net-
work management practices that can affect its quality;
CHAPTER III: THE PROVISION OF CONNECTION AND INTER-
NET APPLICATIONS SECTION I THE NETWORK NEUTRALITY
Article 9. The agent in charge of transmission, switching or routing is
obliged to treat any data package with isonomy, regardless of con-
tent, origin and destination, service, terminal or application.
§ 1st - The discrimination or degradation of traffic will be regulated in
accordance to the private assignments of the president of the repub-
lic provided in item IV of the article 84 of the constitution, to the
faithful implementation of this law, being heard the Internet steering
80 In the Us, the public debate about this topic, between Tim Wu (advocate of the
neutrality) and Christopher Yoo (against it), was extremely relevant for the matter:
T. WU, C.S. YOO, Keeping the Internet Neutral? Tim Wu and Christopher Yoo Debate,
in Federal Communications Law Journal, v. 59, n. 03, p. 575-592, June 2007.
81 BRAZIL, National Congress. Law n. 9.472. Brasília: 16 July 1997. “Article 3. The
user of the the communication services have the right to:
(…)
III - not to be discriminated about the conditions of access and usage of the ser-
vices”.
LEONARDO PARENTONI
132
committee (CGI) and the national agency of telecommunications
(ANATEL), and may only arise from:
I - technical requirements essential for the adequate provision of ser-
vices and applications; and
II - emergency services prioritization.
§ 2nd - In the event of discrimination or degradation of traffic referred
to in paragraph 1, the aforementioned agent must:
I - to refrain from causing damage to users, as regarded in article 927 of
the civil code;
II - to act with proportionality, transparency and equal protection;
III - to inform previously the users in a transparent, clear and sufficient-
ly descriptive manner about its management and traffic mitigation
adopted practices, including those related to network security; and
IV - to provide services on non-discriminatory commercial conditions
and refrain from practicing anticompetitive behaviors.
§ 3rd - In the provision of Internet connection, onerous or for free, as
well as in the transmission, switching or routing, it is forbidden to
block, monitor, filter or analyze the contents of data packets, re-
spected the provisions of this article82.
Network neutrality was one of the most controversial points of the
Brazilian Civil Rights Framework for the Internet, and it was discussed
throughout the whole legislative process83. If the interception of elec-
tronic communications scandal in several States by the Government of
the United States (known as case Edward Snowden)84 haven’t hap-
pened, perhaps today the Brazilian Civil Rights Framework for the In-
ternet would not have been approved in Congress. Mainly because the
priority, in the following years, was the serious economic and political
crisis that took place in the country.
In general, the Brazilian legal literature has proven favourable to the
Brazilian Civil Rights Framework for the Internet, although there are,
of course, some critics85 to the way the network neutrality was treated.
82 BRAZIL, National Congress. Law n. 12.965. Brasília: 23 April 2014.
83 For a detail history of this topic: R. LEMOS, op. cit.
84 Open content. Wikipédia: the free encyclopedia. “Edward Snowden”, available at
<https://pt.wikipedia.org/wiki/Edward_Snowden>, access 04 April 2016.
85 Some examples of harsh critics: E.T. FILHO, O Marco Civil da Internet e as Li-
berdades de Mercado, in N. DE LUCCA, A. SIMÃO FILHO, C.R.P. DE LIMA (Coord.),
Direito & Internet III: Marco Civil da Internet - Lei nº 12.965/2014, São Paulo, 2015,
t. II, p. 59-60.
NETWORK NEUTRALITY
133
It is understood that this law – despite it being timely and relevant – is
not an end in itself. It represents just another step towards the legal reg-
ulation of the subject. In order to reach its objectives, the net neutrality,
according to the view of this author, must: 1) be contextualized accord-
ing to the original design of the Internet and its infrastructure; and
2) have clear limits.
It is worth noticing, besides, that almost in the end of her govern-
ment, less than 24 hours before being removed because of the im-
peachment proceedings, the President Dilma Rousseff edited a Decree86
regulating the Brazilian Civil Rights Framework for the Internet, refer-
ring to two controversial issues: 1) net neutrality; and 2) the processing
of personal data.
Despite the possible censorship referring to this moment – political
and legal – based on which this Decree was published, the rules that it
contains, generally, are in accordance with the ideas defended in this
essay and can be considered a breakthrough in the topic.
In the next topic, these issues will be best addressed.
“In theory, this neutrality claim may be beneficial to Internet users in Brazil, but it
also may have little practical effect because, because if the network is global, with data
traffic from one point to another of the world, the neutrality would not have much use if
other countries have not required the neutrality of their networks in their territories. It
does not seem possible for the data to flow in Brazil with neutrality if the same data is
transferred in different conditions in the networks of other countries.
(…)
It should be noticed that the Brazilian Civil Rights Framework for the Internet, did
not consider the regulations that already existed in the General Telecommunications
Law (Law 9472 of July 16, 1997). The art. 3 of that law already guaranteed users of
telecommunications, among other rights, ‘III - not to be discriminated against on the
conditions of access and enjoyment of the service’ (…).
Thus, any anti-competitive or discriminatory practices coming from the Internet
connection providers, which could adversely affect the market freedom, did not result
of the absence of applicable legislation, but rather of the inefficiency in combating such
practices. Besides, the regulation of the General Telecommunications Law is more
precise than the the one of the Brazilian Civil Rights Framework for the Internet, in
which they used intentionally vague terms and terms with little semantic content, pre-
cisely to facilitate its adoption as an ordinary law by the National Congress”.
86 BRAZIL, President of the Republic. Decree n. 8.771. Brasília: 11 May 2016.
LEONARDO PARENTONI
134
7. Internet design and the limits of network neutrality
Both in Brazil87 and in the world88, the debate over network neutrali-
ty is highly polarized. On one side, there are those that oppose to it.
They argue, essentially, that the absolute neutrality runs counter to the
basic notion of market because it eliminates the competitive ad-
vantages, discouraging investment in research and innovation. After all,
what motivates innovation is exactly the temporary monopoly granted
by the intellectual property system, so the creator can refund the costs
that he had and still make profit. If he does not receive differential
treatment to their products, services or processes, rather than the com-
petitors, there will be no incentive to innovate. Those who share this
opinion also argue that the differential treatment to certain data packets
is inner to the management of computer networks, in a way that a fully
neutral network would have low quality of services and which could
lead to the dissatisfaction of the users. Certain controls are necessary,
for example, as in the packet filtering in order to avoid spam89, or when
it seeks to prevent that only a few users (heavy users) consume all the
connection90. These and other arguments are typical of Internet access
providers and large companies in the telecommunications industry. Ul-
timately, these agents intend to ensure full freedom to administer the
connection of the users, as they see it is the best.
87 P.H.S. RAMOS, op. cit., p. 152. “The issue of net neutrality often brings ideologi-
cal views and discourse that has little to do with scientific research. Business, political
activists, lobbyists and lawmakers, although playing valid roles within the political
debate, often carry his speeches with shallow arguments and little empirical evidence
that, in no way, contribute to go deeper in the debate on the effects of regulation of net
neutrality for the society”.
88 For both arguments in favor of and against net neutrality, see: T.M. LENARD,
R.J. MAY (Coord.), Net Neutrality or Net Neutering: Should Broadband Internet Ser-
vices Be Regulated?, New York, 2006.
89 L. PARENTONI, SPAM: presente, passado e futuro, in Revista de Direito das Co-
municações, ano 3, n. 5, p. 13-48, jan./jun. 2012. “(…) Spam is the electronic message
that has direct or indirect commercial nature, and that is massively sent to multiple
receptors, consciously, with uniform content and without any potential interest to the
receptor”.
90 The excessive use of peer-to-peer applications, for example.
NETWORK NEUTRALITY
135
At the other extreme are the advocates of the network neutrality.
Generally, some governments, non-governmental organizations, some
engaged Internet users and some academics. They argue that net neu-
trality is consistent with the original design of the Internet, being large-
ly responsible for the World Wide Web to become the success it is to-
day. Furthermore, they argue that this design is vital to maintain the
collaborative nature and innovative potential of the Internet. Thus, any
filtering of the data packets would be, at first, harmful. Even those fil-
tering that are based on security reasons or in the network management
should be justified and adopted only in exceptional cases. In addition,
the packet filtering could lead to censorship or illegally restriction of
the user’s choices, as in the case of traffic shaping. Therefore, the ones
in favour believe that the scope of the net neutrality should be expanded
as much as possible.
This essay is not linked to any of these extremes. In fact, it considers
that both bring relevant arguments, which must be taken into considera-
tion in order to understand the limits of network neutrality. It seeks, in
this context, an intermediate route, which recognizes the need to pre-
serve the competitive advantages of the entrepreneurs (whatever their
economic size might be) and, at the same time, to allow independent
developers and ordinary people to use the Internet as a tool for innova-
tion and for sharing content. In this sense, what is proposed is a legal
regulation that first understands the functioning of the Internet and,
only then, will set limits to the network neutrality, in accordance with
the structural changes experienced in recent years.
But if the original Internet design was based on certain principles, as
discussed in this text, it should also occur with the legal regulation.
Two legal principles are worth mentioning: 1) The end-to-end commu-
nication; and 2) The preservation of the network layers (layers princi-
ple).
From the computer network architecture’s point of view, the end-to-
end principle is already old and, according to that, the lower layers of
the network must be designed in a way that it can work in the simplest
way possible, getting the complexity reserved for upper layer applica-
LEONARDO PARENTONI
136
tions91. This way, many different applications, with different degrees of
complexity, are able to coexist in a way that all of them are benefited
from the same standard infrastructure of data, which requires fewer
changes in the network, making it more stable. In the case of the Inter-
net, this means that the TCP/IP protocols should have as fewer inter-
ventions as possible. Innovation should preferably occur in the upper
layer, with the development of new applications.
Bringing this topic to the legal analysis, decades later, this principle
states that the legal regulation of the Internet should focus on the appli-
cation layer, preserving, as much as possible, the original design of the
other network layers92. The Napster case, already mentioned, is a good
example. Focusing exclusively on the Napster application, it was possi-
ble to combat the unlawful conduct, without changing the structure of
the Internet or compromising the P2P technology. Thereof, the applica-
tions that emerged later made use of P2P to provide new services. On
the other hand, the Chinese model makes usage of the TCP/IP protocols
for political censorship, which would be prohibited under this principle.
Some authors divide this principle in two different aspects. Accord-
ing to the broad version93, the end-to-end principle determines that the
lower network layers must provide increasingly standardized and sim-
plified services. For example, the physical layer (connecting cables)
91 J.H. SALTZER, D.P. REED, D.D. CLARK, End-to-End Arguments in System Design.
ACM Transactions in Computer Systems, New York, v. 02, n. 04, p. 277-288,
Nov. 1984.
92 L. LESSIG, Code: Version 2.0, cit., p. 44. “Rather than build into this network a
complex set of functionality thought to be needed by every single application, this net-
work philosophy pushes complexity to the edge of the network – to the applications that
run on the network, rather than the network’s core. The core is kept as simple as possi-
ble”.
93 B. VAN SCHEWICK, op. cit., p. 96. “The broad version of the end-to-end argu-
ments argues that application-specic functionality usually cannot – and preferably
should not – be implemented in the lower layers of the network, the network’s core.
Instead, a function should be implemented in a network layer only if it can be complete-
ly and correctly implemented at that layer and is used by all clients of that layer. Thus,
lower layers, or the core of the network, should provide only general services of broad
utility across applications, whereas application-specic functionality should be imple-
mented in higher layers at the end hosts”.
NETWORK NEUTRALITY
137
takes care exclusively of the transport of the data in its raw form. The
way to connect to it is a standard. Differently, the upper layer, the ap-
plication one, performs various activities. In this layer, are performed
simple and complex tasks, involving the processing of data in order to
generate results as text, images, videos, etc. Consequently, any changes
in the network layer should only be made if they can be fully imple-
mented in this layer, without it meaning that the other layers have to be
changed too. And only when these changes are essential for all services
that use this layer. For example, there is no justification to change the
TCP/IP protocols to improve the quality of a specific application, be-
cause it is not essential for the operation of the others. This attitude
would lead to the disruption of the pattern in the lower layers, adding
them unnecessary complexity. Moreover, this kind of change can com-
promise the simplicity and stability of the network. In the other hand,
the replacement of copper cables for the fiber optic ones is a change
that can be implemented entirely in the physical layer, benefitting all
the users, without increasing the complexity in this layer. This substitu-
tion would, therefore, be in accordance with the principle in analysis.
The narrow version94 of this principle provides that all changes that
cannot be implemented solely in one network layer, or that would not
beneficiate all the services that make usage of this layer, without any
distinction, must be implemented in the application layer, functioning
end-to-end, exclusively between the sender and receiver. Thereof, if
any application is developed and it requires a functionality that still
doesn’t exist (such as a new type of biometric identification), it is nec-
essary that all the technology required for its operation is inside the ap-
plication itself. This way, it is enough that the users only have this ap-
plication installed on their devices so they can communicate, without
the need for any change in the Internet infrastructure.
94 B. VAN SCHEWICK, op. cit., p. 90. “The narrow version of the end-to-end argu-
ments postulates that, since certain functions cannot be completely and correctly im-
plemented in lower layers of the network, they must be implemented end-to-end be-
tween the end hosts that are the original source and the ultimate destination of data”.
LEONARDO PARENTONI
138
Another important legal principle is the preservation of network lay-
ers (layers principle)95. According to this principle, only in exceptional
cases the regulation of one layer can interfere with the operation of the
others. Returning to the case of China, the solution adopted by the gov-
ernment is contrary to this principle because, to combat a problem in
the application layer (access to political content), they reached central
layers of the network (TCP/IP). An extreme example would be a coun-
try that, intending to exercise the maximum control over the Internet,
would keep a single access provider, which would be tightly supervised
by the State. In this case, instead of focusing only on the application
layer, the control is focusing on the connection layer.
The regulation that goes further of the network layer brings two
problems that are apparently antagonistic: over inclusiveness and under
inclusiveness. Once more referring to the case of China, it is forbidden
access to any sites that contain unofficial political content. Neverthe-
less, many of these sites also provide other information on various sub-
jects, which there is no need to be blocked (excessive coverage). On the
other hand, sites that do not directly are dedicated to political discus-
sion, doing it in an indirect way, would be immune to the block (insuf-
ficient coverage). Considering all that, the best legal regulation is the
one that is directed exclusively to the layer where the problem is – usu-
ally the applications layer – without any effect on the others.
Both end-to-end and layers principles are the basis of the network
neutrality. It is true that the TCP/IP protocols have undergone profound
changes in recent years. Its original neutrality has been replaced by a
95 L.B. SOLUM, M. CHUNG, op. cit., p. 817-818. “Our thesis is that the design of le-
gal rules should respect a fundamental principle of Internet architecture, which we shall
call the layers principle. At this stage, we can roughly formulate the layers principle as
a rule of thumb for Internet regulators: respect the integrity of the layers. This funda-
mental principle has two corollaries. The first corollary is the principle of layer separa-
tion: Internet regulation should not violate or compromise the separation between layers
designed into the basic architecture of the Internet. The second corollary is the principle
of minimizing layer crossing: minimize the distance between the layer at which the law
aims to produce an effect and the layer directly affected by legal regulation. (…) The
best regulations attack a problem at a given layer with a regulation at that layer. The
worst regulations attack a problem at the content layer by imposing a regulation at the
physical layer – or vice versa”.
NETWORK NEUTRALITY
139
detailed investigation of the packages. Thereof, it becomes necessary to
set legal limits to avoid the negative consequences of such monitoring.
These limits are given precisely by the network neutrality. Thus, it is
necessary to systematize it.
The network neutrality – as any other principle – is not an absolute
value96. It should be relativized considering certain legal reasons or
factual characteristics of each network.
Thereof, some factual characteristics of the Internet require special
treatment for certain data packets, in a way that the final result of the
process is satisfactory97. For example, the differential treatment is justi-
fied, having in mind the nature of the e-mail, in comparison with the
streaming of video in real time. The delay of a few seconds to deliver
an e-mail does not compromise the functionality of this application. In
fact, it is not even noticed by the receiver. In the other hand, constant
delays in displaying a video, even for only a few seconds, harms the
functionality of this application. From the factual and technical-opera-
tional points of view, therefore, the different treatment of these data
packets is justified, reserving higher transmission speed to streaming
video, compared to email. The Brazilian Civil Rights Framework for
the Internet expressly admits this type of discrimination98, only for
technical reasons, for the sake of a higher quality of services (network
management)99. The Decree nº 8.771/2016, that regulates the Brazilian
96 Remember that, in Brazil, even life is not an absolute value, because even the
death penalty is possible in case of war formally declared (CF/88 art. 5º, XLVII, “a”).
97 T. WU, C.S. YOO, op. cit., p. 577. “Yet I don’t think that the fact that an absolute
ban on discrimination would be ridiculous undermines the case for discrimination laws.
It’s like what nutritionists say about fat: there are good and bad types. And what I think
is going on in the network neutrality debate – the useful part of it – is getting a better
grip on what amounts to good and bad forms of discrimination on information net-
works”.
98 BRAZIL, Nacional Congress. Law n. 12.965. Brasília: 23 April 2014. Article 9º,
§ 1º, I.
99 BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. “Article
6. For proper delivery of services and applications on the Internet, the network man-
agement is allowed in order to preserve its stability, security and functionality, using
only technical measures compatible with international standards, developed for the
LEONARDO PARENTONI
140
Civil Rights Framework for the Internet, explains how these limitations
can and should occur100. It starts by pointing that these limitations only
affect the “connection providers and Internet applications”. They do not
reach, for example, “telecommunications services that are not intended
for providing Internet connection” or those “intended for specific
groups of users with strict control of admission”, such as private net-
works (intranets)101.
Another cause for lawful factual discrimination in the data flow is
the emergency services102. It is even recommended to prioritize data
packets that concern them, such as police or public health information,
giving them greater speed of transmission. Which ones specifically are
these services is a matter that should be included in regulations to be
issued by the National Telecommunications Agency - ANATEL103.
In more exceptional situations, such as a formally declared war, it
would be lawful even to interfere with the physical layer and the con-
nection in order to difficult or suppress the access to the Internet from
the belligerent State. If, as mentioned, even the death penalty is permis-
proper functioning of the Internet, and in compliance with regulatory standards issued
by Anatel and considered the guidelines established by CGIbr”.
100 BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. “Article
1 - This Decree points the admitted discrimination of data packets on the Internet and
the traffic degradation; it indicates procedures for data storage and protection by pro-
viders of connection and applications; it points transparency measures on the request of
information for registration by the public administration; and it establishes parameters
for surveillance and investigation of violations of the Law 12.965 of 23 April 2014”.
“Article 4. The discrimination or the degradation of traffic are exceptional
measures, to the extent that it can only come from necessary technical requirements for
the proper provision of services and applications or prioritization of emergency ser-
vices, requiring the obedience to all the requirements set out in article 9, § 2, of Law
n. 12.965/2014”.
101 BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. Arti-
cle 2.
102 BRAZIL, Nacional Congress. Law n.12.965. Brasília: 23 April 2014. Article 9º,
§ 1º, II.
103 BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. Arti-
cle 8.
NETWORK NEUTRALITY
141
sible in this context, even more restrictions on the Internet are al-
lowed104.
Anyway, even where technical and operational reasons justify the
differential treatment of data, it is the network administrator’s duty to
inform the users which are the criteria used and how the various kinds
of data will be handled105. Thus, the users would be able to question
any positions that might lack in technical basis or that might be dispro-
portionate.
These and other exceptions lead to the conclusion that despite inter-
nationally acclaimed, the term “network neutrality” may not be the
most appropriate. After all, the Internet nowadays is not neutral. Some
distinctions are absolutely necessary and legally permitted. What this
principle seeks, in fact, is to prevent and stop unfair discrimination.
Thereof, the network neutrality prohibits the discriminatory treatment
of the same type of data packets, based on their content, origin or desti-
nation106. For example, the distinction between two films, simply be-
104 L.F.T. VERGUEIRO, op. cit., p. 633-634. “Sovereignty means that the State can
control the entire infrastructure from cyber activities that develop the onshore portion of
its territory, in its internal waters, territorial sea (including seabed and subsoil), archi-
pelagos and the underlying airspace. As a consequence of state sovereignty, the cyber
infrastructure should be submitted to the legislative and regulatory control of the state,
but also the State should protect this infrastructure, if it is private or public property.
(…) In the cyber context, the principle of sovereignty allows a state to restrict or pro-
tect, in whole or in part, the access to the Internet, considering the other rules of Inter-
national Law and Human Rights. (…) Similarly, the sovereignty exercised by a State on
its territorial sea grants the full control over the laying of submarine cables, which is a
crucial factor of control, to the extent that the portion with more massive Internet data
currently is supported by submarine cables”.
105 BRAZIL, Nacional Congress. Law n. 12.965. Brasília: 23 April 2014. Article 9º,
§ 2º, III.
BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. Article 5
and article 7.
106 P.A. FORGIONI, Y.R. MIURA, O Princípio da Neutralidade de Rede e o Marco
Civil da Internet no Brasil, in N. DE LUCCA, A. SIMÃO FILHO, C.R.P. DE LIMA (Coord.),
Direito & Internet III: Marco Civil da Internet - Lei nº 12.965/2014, São Paulo, 2015,
t. II, p. 113-114. “The discussions on the neutrality and the equal treatment of data must
take into account that, in the current web architecture, there is already a differentiation
of the information that goes trough its structure: equality is applied to identical or simi-
LEONARDO PARENTONI
142
cause one is from Netflix and other from Net Now. The distinctions can
be made in many ways, not only by changing the speed of access to any
of these applications, but also, e.g. securing that access to one of them
is “free” (in the sense that it will not be deducted from the user’s data
allowance), while access to others will be normally charged.
Thus, it is fully acceptable that the service providers offer various
service packages with different characteristics in each of them. For ex-
ample, services with lower speed and lower franchise data, cheaper, and
services with higher speed and higher deductibles data (or even no de-
ductible), with a higher cost. This is inherent in the market economy.
Stratification of a particular service in different levels, with their own
characteristics and different costs, is common in many areas, even in
the health business, where many different plans could be offered.
Moreover, this differentiation is benefic, since each person or group
have specific needs and capabilities in terms of payment, and it is not
reasonable nor efficient to impose just one model for all.
What is not admitted is that the network administrator, outside the
legally permissible exceptions, would handle the connection of its us-
ers, having influence, even indirectly, in the way that the service pack-
age is used. The choice of how to use the Internet access is personal,
being done by the users. The ISPs cannot provide the connection in or-
der to direct users to specific applications. This rule, which could al-
ready be perfectly understood, in a systematic interpretation of the Bra-
zilian Civil Rights Framework for the Internet, now is properly regulat-
ed107. Once the connection speed and other technical aspects remain the
lar data, and not to the different applications which, because of its characteristic, require
different treatment”.
107 BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. “Article
3. The requirement of an equal treatment brought by article 9 of the Law
n. 12.965/2014, must ensure the preservation of the public and unrestricted nature of the
Internet and the foundations, principles and objectives of Internet use in the country as
provided by the Law n. 12.965/2014”.
“Article 9. Unilateral conducts or agreements between the responsible for transmit-
ting, switching or routing and application provider are forbidden if:
I - commit the public and unrestricted nature of the Internet and the foundations,
principles and objectives of Internet use in the country;
II - prioritize data packets due to commercial arrangements; or
NETWORK NEUTRALITY
143
same, it is the user himself who will decide whether to use Netflix or
Net Now, WhatsApp or any other similar application.
In addition, the network neutrality focuses on the lower layers, espe-
cially the TCP/IP protocols, preserving its full compatibility with any
applications that already exist or will exist. In the application layer, the
rule is the free competition. Each developer can and should create in-
novative technologies, increasingly complex and specific, exploring
them economic and exclusively within the limits of law. In other words,
innovation flourishes in the application layer, from end to end, while
the network neutrality focuses on the lower layers, ensuring that they
are not manipulated to seek unlawful goals.
This way, it is understood that free competition, the business strate-
gies and the competitive advantages can be ensured, on one side, and
privacy, freedom of speech, innovation and the decision-making auton-
omy of users, on the other.
At this point, it is necessary to question the article 2, caput, of De-
cree nº 8.771/2016108, since it also extends the neutrality to the provid-
ers of “Internet applications”, i.e., the upper layer. This article suffers
from two vices. The first is its illegality, since it goes further from the
Brazilian Civil Rights Framework for the Internet, increasing restriction
that was not expressly provided by the law. After all, the article 9 of the
Brazilian Civil Rights Framework for the Internet is clear in providing
that net neutrality applies to the “responsible for transmitting, switching
or routing” data. In other words, it applies to the administrators of the
connection, that operate the lower layers of the network, such as
TCP/IP. There is no nothing in the law that provides that the neutrality
III - prioritize applications offered by themselves, that is also responsible for trans-
mitting, switching or routing or by companies that are member of the same economic
group”.
“Art. 10. The commercial offers and Internet access charging models should pre-
serve a single Internet, with open nature, plural and diverse, seen as a way to promote
human, economic, social and cultural development, contributing to form an inclusive
and non-discriminatory society”.
108 BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. “Article
2. The provisions of this Decree are intended to the companies responsible to transmit,
switch or route and to the connection and Internet application providers, as defined in
the item I of the caput of article 5 of the Law n. 12.965/2014”.
LEONARDO PARENTONI
144
would also reach the application layer. There are several reasons for
this, as analyzed in this essay. The fact that the article 9 is included in a
Chapter entitled “The provision of connection and Internet Applica-
tions” does not change this aspect. Thus, it is understood that the net-
work neutrality is directed specifically to the connection providers,
while the other provisions of this chapter as the “Protection of Records,
the Personal Data and Private Communications”, also go to application
providers. From this point of view, the article 2, caput, of Decree nº
8.771/2016 goes further from the limits set by the law that it intended to
regulate.
This article also suffers from a second vice. Thereof, extending the
network neutrality also to application providers, that operate exclusive-
ly in the top layer of the Internet, conflicts with the constitutional prin-
ciples of free enterprise and free competition. As stated in the text, in
this layer the rule is that every entrepreneur can benefit from competi-
tive advantages, as a result of the innovations he has caused within the
limits of legality. It is precisely the temporary privilege of exploitation
of intellectual creations, exclusively (as in patents and software), which
justifies and stimulates the innovation by encouraging investment in
research and development areas. Stop this can cause serious side ef-
fects.
For these reasons, and in accordance with all the ideas developed
throughout this text, it is reiterated that the network neutrality focuses
on the lower layers, especially the TCP/IP protocols, preserving its full
compatibility with any applications that already exist or that will exist.
In the application layer, differently, the rule is free competition. Each
developer can explore exclusively within the limits of law, the intellec-
tual creations there were developed. Thus, net neutrality is the prohibi-
tion to network administrators to manipulate the connection of users in
order to discriminate the same type of data packets, based on their con-
tent, origin or destination, outside the legally permitted exceptions.
NETWORK NEUTRALITY
145
8. Conclusion
The Internet is not the work of nature but the result of a human con-
struction. It arose as a result of a unique historical moment, which
brought together the military interest of the US Government, concerned
with national security, and the participation of the academic elite of the
country, centered on the ideal of freedom and sharing ideas. With these
features it has consolidated and evolved, becoming one of the most im-
portant inventions of the recent history of mankind.
However, the Internet has changed considerably in recent years.
Some of its main features were replaced, to allow greater monitoring
and control. The TCI/IP protocols, which were formerly neutrally used,
now inspects the data packets, differentiating them by the source, desti-
nation or transmitted content. These changes opened the door for many
forms of censorship and manipulation of users, which was not even im-
agined in the beginning of the Internet.
Such changes have attracted attention from various fields of science,
including the law. This paper has tried to contextualize, define and sys-
tematize the one that, in the author’s view, is the main legal response to
new challenges: network neutrality. Maybe with its correct application,
it might be able to recover the balance of forces that exited in the ori-
gins of the Internet.
ResearchGate has not been able to resolve any citations for this publication.
Law n.12.965. Brasília: 23
  • Nacional Brazil
  • Congress
BRAZIL, Nacional Congress. Law n.12.965. Brasília: 23 April 2014. Article 9º, § 1º, II.
  • President Brazil
  • Of Republic
BRAZIL, President of Republic. Decree n. 8.771. Brasília: 11 May 2016. Article 8.