CSTI Talks
CSTI SEMINAR ON 2
DECEMBER 2002
The CSTI has provided this report for information
purposes and to stimulate and guide thought about the issues involved.
It does not reflect the opinion of the CSTI or of any of its members
or of any of the participants in the seminar.
Download the document in pdf format
PROGRAMME FOR ROUND
TABLES
INTRODUCTION :
Presentation of the seminar and the issues
ROUND TABLE N°
1
Unbundling: incentives and disincentives for high-speed
access
ROUND TABLE N°
2
Digitisation and digital rights: who pays what?
ROUND TABLE N°
3
Voice and image on IP: technological convergence
ROUND TABLE N°
4
Bringing
Europe out of the telecoms crisis
ROUND TABLE N°
5
Critical technologies: perception and reality
ROUND TABLE N°
6
Confidence
and usage: perception and reality
INTRODUCTION :
Presentation of the CSTI
seminar on 2 December 2002
The role of the CSTI is to present the Prime
Minister with guidelines, opinions and initiatives on a certain
number of matters relating to the development of information technologies
in France. Created a little over a year ago, its working methods
are now established. A number of opinions have been published,
some of them followed by action, as recent events have shown.
Many of the most representative figures from
the IT world have gathered here today, 2 December 2002, including
business leaders from large groups to start-ups, researchers and
financiers. The meeting is not open to the public and its aim
is to promote contacts between existing networks in the ICT sectors.
An initial public meeting took place at the Ministry for Research
in March.
The members of the CSTI wanted this meeting
to take the form of a seminar. The discussions will be anonymous,
and the CSTI will not be issuing any opinions based on the conclusions.
The issues
After several years of double-digit growth,
players in the IT industry are now experiencing a brutal reversal
of fortunes. Their financial situation has deteriorated dramatically
over the last two years, leading to far-reaching restructuring.
At the same time, ICT use is continuing to grow:
the number of internet and mobile phone subscribers has risen
sharply and there has been a rapid expansion of high-speed internet
access. Despite a rash of dot.com failures, e-commerce results
in 2002 hold out the hope of a stable market and a long-term future
for the players that remain.
The question now facing them is how to revive
a sector not only in financial difficulty - as can be seen from
the debt levels of certain operators - but also prey to strategic
doubt. The example of WAP shows how difficult it can be for operators
and service providers to anticipate customers' needs and expectations.
The government has a number of means at its
disposal for aiding recovery in the ICT sectors. The measures
may be applied to different parts of the value chain, from the
equipment manufacturer supplying technology to the end-user, via
the operator, infrastructure provider and content provider.
- The legal framework of the digital economy
:
The measures taken to adapt the existing
regulatory framework to digital technologies have been relatively
straightforward, since it is generally able to deal with the
questions that emerge. However, some specific issues have arisen,
mostly relating to intellectual property (especially copying
for private use) and the protection of personal information.
EU directives in both areas are being transposed
into French law, giving rise to heated debate.
One round table considered digitisation
and digital rights in the light of the underlying question
of content providers' business models. Another round table looked
at the question of personal information and, more generally
the technical, legal and regulatory measures that are needed
to reassure consumers and instil a climate of
confidence. Particular attention was paid to the link between
usage and secure transmission, or, more accurately, consumer
perception of levels of security.
- Regulation of the telecommunications market
that allows new players to enter and ensures fair and effective
competition :
At the same time, the regulatory framework of the French telecommunications
industry is due to be overhauled in 2003, with the transposition
of a set of EU directives commonly called the "telecoms package".
The new framework should lay the foundations for harmonisation
of the rules governing electronic communications and pave the
way for convergence in IT, telecommunications and audiovisual
technologies and applications. A round table focused on the reality
of this convergence.
Regulators must aim to ensure effective, long-term
competition on the relatively undisputed telecommunications
market. Although lower interconnection charges have led to a
substantial fall in telephone charges and the cost of internet
access via the conventional telephone network, the rapid spread
of high-speed access raises the specific issue of access charges
and of unbundling in particular. Although the cut in unbundling
charges in April 2002 led to a rapid fall in prices, with a
proliferation of offers at around 30 euros, the question remains
as to the effectiveness of unbundling in
terms of encouraging the spread of DSL technology. One of the
round tables looked at this issue.
- Support for R&D :
In the current economic climate, and with their cash-flow being
squeezed, operators and equipment manufacturers are drastically
cutting their R&D spending. A round table was asked to look
at critical technologies and the resources
that need to be devoted to them so that France and Europe can
stake out advantageous positions.
Another round table reviewed the action that
needs to be taken in order to take Europe out
of the telecommunications crisis, including an in-depth
look at the causes of the crisis. It dealt as a priority with
the situation of equipment manufacturers and operators
ROUND TABLE n° 1
UNBUNDLING : INCENTIVES AND DISINCENTIVES FOR HIGH-SPEED
ACCESS
Presentation by Mr. Laurent Zenou, of the Analysis
consulting firm, of a study
carried out for the CSTI.
Introduction
This presentation describes the main results
of a study
carried out by Analysis for the CSTI. It is an international
comparison of the conditions under which unbundling has been carried
out in countries where high-speed access is relatively
widespread.
Can unbundling drive the spread of high-speed
access ?
The study looked at factors that encourage the
spread of high-speed access, looking in particular for any connection
with unbundling, but did not find any explicit evidence of such
a link. Consider two extreme cases :
- South Korea is famous for the rapid expansion
of high-speed access even though unbundling was not officially
launched until the end of August 2002. The spread of high-speed
access in South Korea is due to a highly proactive policy by the
government, which has sought to encourage the roll-out of high-speed
infrastructure (DSL and cable).
- In Japan, where unbundling is relatively significant
(58% of DSL lines are unbundled at the present time), the spread
of high-speed access has been relatively limited.
Nevertheless, the spread of high-speed access
seems to be linked to alternative operators' development of unbundled
lines, and the time lag between that factor and the incumbent's
launch of DSL services needs to be taken into consideration.
Unbundling: incentives and
disincentives
There are a number of factors that hold back
unbundling :
- Behaviour of the incumbent, which may
sometimes be regarded as discriminatory.
- Access to information : Alternative
operators often find it difficult to get hold of information about
matters such as line quality, frame coverage zones, the characteristics
of colocation facilities, etc. In Germany, for example, information
about the quality and length of lines is sent manually (by post)
to alternative operators, but electronically (ie, much more quickly
and effectively) to subsidiaries or divisions of the incumbent,
which offers high-speed access via DSL.
- Information processing time :
All procedures take longer for alternative operators, who depend
entirely on the incumbent.
- Allocation and operation of colocation facilities
: In the UK, the process initially left considerable freedom
to the players concerned (bow-wave process). However, the system
was inefficient because conflicts of interest soon appeared between
alternative operators, paralysing the process for allocating colocation
facilities.
- Responsibility for responses to operating
incidents : The question has also arisen of who is responsible
for action that needs to be taken to resolve problems or incidents
with line operations. This has been a particularly important factor
in Japan.
- Learning curve : New entrants, who generally
launch their unbundled DSL services later than incumbents, are
lower down the learning curve and consequently lag behind at the
operational level, which impacts on their efficiency in implementing
unbundled lines.
A number of success factors can be identified
from successful international initiatives :
-
A regulator that exercises effective oversight
without excessive intervention. . In Germany, the regulator
acts more as a moderator, leaving alternative operators and
the incumbent to resolve matters between themselves, intervening
only when a real problem arises.
- Rapid and effective regulation : Regulatory bodies
or other authorities (competition authorities, for example)
that have real powers and can respond rapidly (ie, issuing decisions
within three months after a referral) also contribute to success.
-· Long-term visibility: Operators need to be
reassured about future developments in areas such as prices
for totally or partly unbundled lines. Objectives set by the
regulator enable alternative operators to prepare their roll-out
plans more effectively and to be clear about the economic conditions
in which they will be operating.
- Exchanges of information : LExchanges
of information in electronic form generate obvious time savings,
as do automated processes and parallel (rather than sequential)
procedures.
This last factor has a major impact on :
- the availability of colocation facilities, which brings new
zones for unbundling into play;
- - line-by-line delivery, which enables operators to be more
reactive to their customers, a particularly important factor for
business users.
- Co-mingling : Co-mingling is an unbundling
option in which the alternative operator's equipment is installed
in the same facility as the incumbent's hardware, generating savings
of time and money and having an important impact on the progress
of unbundling.
What lessons can France learn ?
DSL services have been available since late
1999, from both the incumbent and alternative operators, the latter
launching DSL services not in totally unbundled form but as a
wholesale offering of DSL traffic carried by the incumbent. Total
unbundling was completed in the third quarter of 2001, relatively
late. But the situation is changing rapidly. High-speed access
is making significant progress. DSL, accounting for over 75% of
high-speed lines, is the main contributory factor. Unbundled lines
represent less than 1% of all DSL lines and, even if the figure
has risen from approx. 700 lines to just over 1,000 lines in two
months, the proportion is still very small.
New offerings have appeared since September
and lower retail prices are due to be proposed shortly, mainly
in the Paris region, for totally unbundled DSL lines. Reported
high-speed and DSL customer bases seem to indicate strong growth:
in November, France Telecom announced that it had more than one
million DSL subscribers.
In addition:
- France Telecom is now providing the available information in
electronic form, which could be useful in setting up the process;
- exchanges of information are being automated, with the introduction
of an unbundling extranet;
- the co-mingling option has been available in France since June
2002.
Conclusion
Unbundling is a relatively complex process which takes time. It
is not a miracle solution that will spark off rapid growth in
high-speed access from one day to the next. There is a learning
curve for everyone involved in unbundling, whether the incumbent,
alternative operators or the regulator.
Discussion
There is relatively little competition on the
fixed telecommunications market, unlike the mobile market. Competition
takes two forms: alternative technologies (cable networks and
radio local loop) and the unbundling of the incumbent's copper
local loop.
The purpose of the round table was to consider
the question of unbundling as a means of bringing high-speed access
to French business and residential users. It is important that
unbundling should continue to be one means among others of achieving
this primary objective.
Orders for unbundled lines have risen sharply
very recently. A few weeks ago, the total of all orders amounted
to barely more than a thousand lines, a figure which is now reached
weekly. This is a drop in the ocean compared with the 33 million
lines that exist altogether, but the movement seems to have started.
Implementing unbundling
in France
A wide-ranging debate has taken place in recent
years, moderated by the telecommunications regulatory authority
(ART), on broad pricing principles for unbundled lines. That debate
is now more or less complete.
Unbundling has now entered its operational phase.
Both France Telecom and the alternative operators need to move
up their respective learning curves. Likewise, the introduction
of preselection for voice services also required a certain lead
time. Generally speaking, the alternative operators have come
to the same conclusion: operational processes cannot be established
instantaneously. New processes have to be introduced gradually.
Although there have been some problems in the past, the alternative
operators say that they are now able to maintain a dialogue with
France Telecom staff on the ground, to unbundle lines and to improve
services. The main stumbling blocks that remain relate to co-mingling
and virtual colocation.
The success of unbundling
will depend on the alternative operators' continuing existence.
The real restrictions as far as unbundling are
concerned are:
- capillarity;
- the necessary investment;
- connection of the distribution frame to a network (backbone).
Unbundling is relevant above all in densely
populated urban areas, where installing the DSLAM (the equipment
installed by the third-party operator in France Telecom's distribution
frame) is inexpensive because there is a nearby operator connection
point. In large cities like Paris, Lyon and Marseille, unbundling
will happen.
The next question is what economic model should
underlie this mechanism, which relies on unbundling and on France
Telecom selling on a service. An operator cannot launch an offering
and hope to achieve critical size unless it goes beyond the customer
base served only by the sites where it has decided to unbundle.
The operator therefore needs to supplement its offering with services
sold on by France Telecom (in the event, option 3). Options 3
and 5, which represent not unbundling but France Telecom's resale
of high-speed lines, supplement the unbundler's initiative on
the sites where it is present.
Unbundling everywhere would not make economic
sense. Different technologies must be used in conjunction in order
to open up high-speed access.
· Although DSL can be rolled out quickly, there are real
investment drawbacks. Unbundling takes place in France Telecom's
distribution frames, ie, at a very low level in the network, but
alternative operators' networks cannot connect all the 12,000
distribution frames in France even though they have a certain
capillarity.
· Other solutions must therefore be used to in conjunction
with unbundling, such as radio local loops or wholesale ADSL from
France Telecom (options 3 and 5). They enable alternative operators
to offer nationwide coverage and high-speed DSL to both business
and residential consumers, supporting roll-out at option 1 level
and going as far as possible down the network.
It is misleading to think that unbundling can
be achieved instantaneously nationwide. That was not the case
with voice services, which were also introduced gradually. Operators
began by interconnecting with operator connection points, for
mere transit. Then, as volume increased, they were progressively
able to finance capillarity as far as the self-routing switch.
The movement was gradual because three price levels existed in
France Telecom's interconnection catalogue:
- a dual transit price, with a single operator connection point
nationally;
- a single transit price for operators with a connection point
in each region of mainland France;
- a price per self-routing switch, much lower down the network
and much more attractive in terms of connection time.
If France Telecom offered an intermediate price
at lower than region level, it would enable operators with networks
offering voice services to use them better. The ART is due to
rule on this question of an intermediate tariff.
De facto, perequation does not enter
the operators' business model: they choose unbundling where the
costs are lowest, which is entirely understandable. In contrast,
France Telecom is subject to perequation, raising a fundamental
question that the ART has to address.
Unbundling provides an answer
to the question of service provision but does not solve the problem
of access, which is also crucial in the least densely populated
areas.
While everyone agrees that the local loop is
the bottleneck - which justifies unbundling per se - the question
of competition on the backbones is more controversial. It exists
in principle because that is the level at which competition in
telecommunications was initiated. In many areas, however, there
is virtually no alternative option for connection from the distribution
frame to a high-speed network. Although optical fibre has been
installed along almost all motorways, waterways and railway lines,
it does not always provide a basis for offering profitable services.
If there is a possibility of operating in a sector, it is essential
not to lose money on routing to the interconnection point with
the provider's own network.
Metropolitan networks are one of the keys to
the development of high-speed access for business and residential
users. However, they are of greater interest to businesses than
to residential users, who can have access through capillary links
(radio or copper local loops). Firms on the North American market
are greatly interested in techniques based on optical trees that
would connect directly to new generation metropolitan networks
better suited to the internet. The cost of optical equipment has
tumbled in the last ten years. At the present time, the only drawback
with optical fibre is the civil engineering cost, though techniques
have been developed for laying optical fibres without digging
trenches, as long as there are pipes.
In France, a substantial number of organisations,
especially large local authorities, have realised that action
at metropolitan level, and not only at the very end of the network,
is a key factor for development.
Metropolitan networks (which pull together local
loops before accessing the core network) have particular features
that may make them the weak link in the high-speed chain. Metropolitan
networks today are all based on Sonet/SDH technologies, which
have the great advantage of being extremely robust (for example,
if optical fibres are cut) but permit no flexibility in bandwidth
allocation and are designed to carry telephone rather than IP
traffic. They therefore pose a problem, whatever the preferred
option for the final link (radio, copper, optical fibre, etc.).
High-speed access for local authorities
As far as coverage is concerned, rather than
seeking incentives, operators seem to be asking not to be obliged
to roll out at a loss. Local authorities are not necessarily interested
in unbundling, but they are interested in high-speed internet
access. They can act as a staging post, enabling operators to
extend investment zones for services that would be sold to third-party
operators not in the form of unbundling but as the sold-on services
referred to earlier (options 3 and 5). Again, a distinction needs
to be drawn between unbundling, as one of a number of means of
ensuring competition, and high-speed access itself, which is the
desired end.
The question arises here of whether local authorities
can:
- allow an operator to use equipment which they may directly or
indirectly have helped to acquire;
- benefit from the fact that this equipment, installed in frames,
is pooled for unbundling purposes.
Alternative solutions
Unbundling is one solution; other solutions,
whether technological or regulatory, could be envisaged.
- Has unbundling a chance of working?
- Is it possible to envisage a structural separation between the
incumbent's network and service?
- Could the local public services model enabling local authorities
to invest in a local operator (possibly with a public-private
shareholder structure) to develop alternative access contribute
to the spread of high-speed access in sparsely populated areas?
- Is the introduction of Ethernet in very densely populated areas
(not far from frames) on totally unbundled lines a mature and
attractive option?
- At what point will firms decide to migrate to optical technologies?
ROUND TABLE n° 2
DIGITISATION AND DIGITAL RIGHTS: WHO PAYS WHAT?
Introduction
The question of digitisation and digital rights
raises the following issues:
- What are the limits on totally free content?
- What does the concept of an identical copy mean in the digital
world?
- What added value does digitisation create and how should it
be shared between the players involved?
A draft bill has been prepared prior to transposition
of the EU Copyright Directive. Interested parties can submit their
comments on the bill until 15 December 2002; it will be reviewed
by the CSPLA (High Council for Literary and Artistic Property)
at a plenary session on 5 December 2002.
The bill contains most of the legislative issues
that the participants will be discussing during the round table.
The key issues concern:
- enforcement (anti-piracy measures);
- intellectual property (adaptation of intellectual property law
to the issues raised by digitisation and the information society).
The question of intellectual property raises
issues that are specific to France, since they are linked to French
copyright law (for example, as regards reform of aspects relating
to collective works and artistic creation by salaried employees).
Other matters have a more specifically American,
EU or international dimension, such as digital rights management
and exceptions (fair use exception, private copy exception and
limitative lists of exceptions in the copyright system).
In France, the CSPLA deals with most of these
issues. The Act of 17 July 2001 extended the benefit of remuneration
for private copies to new categories of works, and the Brun-Buisson
Commission is adapting the terms of the 1985 law on private copy
to digital media and new written, graphic and plastic works.
Discussion
A case in point :
images and the teaching of art history.
In art history, photographs are an essential
aspect of teaching, research and the dissemination of knowledge.
The situation has changed considerably in the
last two years.
· New technologies have had a profound
effect on teaching methods, in France and elsewhere. Hitherto,
slides were used for teaching purposes; now, Americans use digital
photographs stored on educational websites. The programmes set
up by major American institutions like Columbia University and
MIT now give on-line access not only to databases but to courses.
· The EU Copyright Directive mentioned in the introduction,
published two years ago, clearly inhibits the teaching of art
history. Art history teaching used to infringe the law, since
slides and digital photos were taken from works or open-air shots
(of monuments) without remuneration.
· Concurrent developments in the sphere of image rights
make it increasingly difficult to reproduce, free of charge, views
of any work, document or even monument (for example, the image
rights to the building where this meeting is being held belong
to the Senate).
This raises several points.
· There can be no question nowadays of
not taking advantage of the possibilities offered by digitisation,
especially that of giving students access to photos they can view
outside the classroom. Education via the internet is also an extremely
promising opportunity that should be envisaged.
· As far as art history is concerned, the emerging legal
framework will mean that photographs can no longer be shown in
courses, since four persons would have to be remunerated:
- the owner of the work,
- the heirs and assigns, when the work is not in the public domain,
- the photographer of the work,
- the provider of the medium by which the photograph is disseminated.
Assuming 200 hours of courses and 30 to 50 photographs per course,
a reporting system on that basis would clearly be impossible to
implement or finance.
Two legal systems that exist in other countries
would provide a solution to this problem.
· In Switzerland, everything to do with teaching is exempt
from these rights, the classroom being extended to include the
web. Art historians in France would like to move in this direction.
· In the United States, the fair use exception implies
that a person is exempt from the requirement when they use an
image for non-commercial purposes. Digital photos are used in
art history courses at Yale University and all of them, including
photos of contemporary works, are accessible to students by means
of a PIN code and are placed online free of all rights.
French students pay a university registration
fee of about 300 euros. Consequently, departments have very few
resources of their own, especially as the fee is the same whatever
the discipline. Universities cannot afford the rights they will
be asked to pay. Even major public institutions like the Bibliothèque
Nationale and the Louvre cannot offer their works free of rights.
Courses draw on a relatively broad corpus of material. It would
be highly restrictive if teachers were allowed to use only the
collections of French or European institutions in their courses.
Another case in point :
books, from the publishers' viewpoint
Comparing teaching conditions in France and
the United States is unjustified, especially as university education
in the US is fee-paying and there is no comparison with the business
model and financial resources of American universities. American
students use networks and databases to find knowledge. Lexis Westlaw,
a legal database, is accessible online from a computer room; the
well-stocked library next door is empty. However, although students
have free access to databases, the university pays royalties to
beneficiaries.
Digital content will gradually replace physical
media. Music on a compact disc can be copied onto other CDs or
stored on a computer. Paper is destined to disappear. The 25 volumes
of Balzac's Comédie Humaine fit onto a single CD ROM.
The publishing industry continues to cling to
individual rights management. Rights can be remunerated in two
ways:
- through individual management;
- through collective management.
In the publishing industry, publishers have
a contractual right to reproduce and represent authors' work and
they remunerate them in consequence. The book is in principle
the medium through which remuneration is obtained, but it will
be superseded by the databases (CD ROMs, etc.) through which the
knowledge is disseminated.
Remuneration is essential for authors and guarantees
creation. Photocopying has already drawn attention to the question
of reproduction: it took 15 years for legislation to reach the
statute book, in the form of the Act of 3 January 1995 which instituted
a system of legal licenses. A rights body, the Centre Français
du Droit de Copie, used to collect over 15 million euros each
year, which it distributed across the whole spectrum of school
and university publishing. However, the system has not been able
to maintain diversity in scientific publishing. Authors and publishers
alike are deeply concerned about the consequences of digitisation.
In music (where collective rights management
has a longer history), in film and in other areas, rights bodies
like the SACEM, SACD and DAGP have formed Sesame, a kind of one-stop
shop where users can pay a negotiable fee according to their type
of use (fees are not the same for commercial and educational or
research use). Rights bodies and the rightholders they represent
therefore have different schedules of rights and different arrangements
for the payment of rights.
In an individual management tradition where
the publisher is the exclusive rightholder, the private copy principle
is necessary for the promotion of knowledge and individual freedom.
An organisation paying the private copy fee can provide students
with copies of works on the internet.
The Brun-Buisson Commission on private copy
has called for royalties on all CD ROM-type media, like those
that still exist for blank video and audio tapes. The fee will
offset lost revenue from use without purchase. For publishers
the remuneration is fair, not only for authors and their beneficiaries
but also for the entire creative community, since 25% of the total
amount of royalties will be earmarked to support creative artists'
personal education. The aim is to strike a balance between the
respective interests of creative artists and the public. In a
country which has always defended a humanist approach to copyright,
safeguarding remuneration and ensuring a rich and diverse culture,
it is essential to recognise this approach and not to create artificial
exceptions for education or research. The means for gaining access
to culture are not the same in Switzerland or the United States.
Recognising the value of a work is a form of culture and an education
in itself. A minimum level of protection must be maintained so
that creative artists can continue to create.
There are new models for managing intellectual
property. That being said, copyright as conceived in 1957 regards
intellectual property as intangible and is therefore well-suited
for the economy of the future. There is no legal loophole in that
regard. All that is needed is to define the base for assessing
the amount to be paid and to share the proceeds appropriately.
The book is a medium. The fact that text and
image can be stored in databases and reproduced and disseminated
on CD ROM and other media is a technological revolution. The revolution
has already taken place for film, with the licence fee for (public)
broadcast television. Likewise, there are legal licences for broadcasting
music on the radio, set up by rights bodies. That being said,
the problem with legal licences is that they can lead to a sameness
in the choice and selection of broadcast works. On another subject,
it is now possible to subscribe to legislative and legal publications,
their authors being remunerated in proportion to the amount of
fees.
La comparaison des conditions de l’enseignement
en France et aux États-Unis se justifie d’autant moins
que l’enseignement outre Atlantique est payant et que le
modèle économique et les moyens financiers des universités
y sont sans commune mesure. La connaissance des étudiants
américains se base désormais sur les réseaux,
les bases de données. La base de données juridique
Lexis Weslaw est accessible dans une salle d’informatique
; dans la salle voisine, où se trouve une bibliothèque
extrêmement riche, il n’y a plus personne. Néanmoins,
il ne faut pas oublier que si les élèves ont accès
à des bases de données, l’Université
acquitte pour ce faire une redevance à des ayants droits.
Les contenus numériques vont progressivement
remplacer les supports physiques. La musique présente sur
un Compact Disc peut être copiée sur d’autres
Compacts Discs ou être enregistrée sur un ordinateur.
L’écrit papier est aussi apparu très tôt
comme un support appelé à se "dématérialiser".
Les 25 volumes de La Comédie Humaine de Balzac peuvent
être contenus sur un seul CD-ROM.
Le secteur de l’édition continue
à se cramponner à une gestion individuelle de ses
droits. La rémunération des droits peut être
effectuée de deux façons :
- par la gestion dite individuelle;
- par la gestion dite collective.
À ce jour, dans l’édition,
les éditeurs disposent par contrat du droit de reproduire
et de représenter l’œuvre des auteurs, et les
rémunèrent en conséquence. Le livre constitue
a priori le support permettant d’obtenir cette rémunération,
mais celle-ci sera de plus en plus obtenue à partir des
bases de données (CD-ROMs…) qui permettent de diffuser
ces connaissances.
La rémunération est essentielle
pour l’auteur, et garantit la création. La reproduction
des œuvres a déjà été problématique
avec la photocopie; 15 ans ont été nécessaires
avant d’arriver à la loi du 3 janvier 1995, qui a
mis en place un système de licence légale, 100 millions
de francs étant recueillis annuellement par une société
agréée (le Centre Français du Droit de Copie),
et répartis au niveau de l’ensemble des œuvres
scolaires universitaires. Ce système n’a cependant
pas permis de maintenir la diversité dans le domaine de
l’édition scientifique. Les auteurs comme les éditeurs,
sont désormais très attentifs aux conséquences
de la numérisation.
Dans le monde musical, où la gestion
collective des droits est l’objet d’une histoire plus
importante, dans le monde du cinéma et dans d’autres,
des sociétés comme la SACEM, la SACD ou la DAGP
(pour les œuvres graphiques) ont constitué la société
Sésame, sorte de guichet unique permettant déjà
aux utilisateurs de s’acquitter d’une redevance négociable
en fonction de l’usage (les rémunérations ne
sont pas les mêmes dans le cadre d’une exploitation
commerciale et dans celui d’une exploitation éducative
ou de recherche). Il existe donc un aménagement des droits
et de l’acquittement des droits par les sociétés
ou les titulaires les représentant.
Dans une tradition de gestion individuelle et
d’exclusivité du titulaire qu’est l’éditeur,
le principe de la copie privée constitue une respiration
nécessaire à la connaissance et à la liberté
individuelle. La rémunération pour copie privée
permet à l’organisme qui s’en acquitte de proposer
à des étudiants des copies sur Internet.
La commission Copie Privée présidée
par M. Brun-Buisson a prévu une redevance sur tous les
supports de type CD-ROM, à l’instar de celle qui continue
d’exister pour les cassettes audio et vidéo vierges.
Cette rémunération permettra de compenser des pertes
de revenu découlant d’un usage hors achat. Pour les
éditeurs, il s’agit là d’une rémunération
équitable, tant pour les auteurs que pour leurs ayant droits,
mais également pour l’ensemble de la communauté
des créateurs puisque 25 % du montant total de la rémunération
aidera les créateurs pour leur formation personnelle. L’objectif
est de trouver un équilibre entre les intérêts
respectifs des créateurs et du public. Dans un pays qui
a toujours défendu un droit d’auteur humaniste, qui
permet une rémunération et assure la diversité
et la richesse de la culture, il semble essentiel que cette approche
soit reconnue, et que ne soient pas créées artificiellement
d’exceptions dans les domaines de l’éducation
ou de la recherche. En Suisse ou aux États-Unis, les moyens
d’accéder à la culture ne sont pas les mêmes.
La valorisation d’une œuvre est déjà une
forme culturelle et une éducation en soi. Il faut maintenir
un seuil minimum de protection pour que les créateurs puissent
créer.
Il existe donc de nouveaux modèles de
gestion de la propriété intellectuelle. Cela étant,
le droit d’auteur, tel qu’il a été conçu
en 1957, considère la propriété intellectuelle
comme une propriété immatérielle, et s’adapte
donc parfaitement à l’économie de demain. Il
n’y a pas carence de droit en la matière ; il est
seulement nécessaire de trouver l’assiette du montant
du droit, et de bien le répartir.
Le livre est un support. Que le texte et l’image
soient intégrés à des bases de données
et soient reproduits sur des CD-ROMs et d’autres supports
qui peuvent le diffuser représente une révolution
technologique. Cette révolution est déjà
une réalité pour le film, puisque la télévision
hertzienne donne seulement lieu à une redevance (qui en
outre ne concerne que le service public). De même, il existe
des systèmes de licence légale pour la diffusion
de la musique à la radio, mis en place par la gestion collective.
Cela étant, le problème de la licence légale
est qu’elle peut parvenir à une unicité du
choix et de la sélection des œuvres diffusées.
Au demeurant, on peut aujourd'hui accéder, via des abonnements,
à des éditions législatives et juridiques,
qui donnent lieu à une rémunération des auteurs
au prorata du montant de droits acquittés.
Copyright at present remunerates not
the author but the publisher, not the content but the medium.
The challenge is for Europe to achieve economic
leadership within the next few years (in accordance with the Lisbon
criteria) in a new economy in which value creation is based not
on transactions but on sharing (sharing knowledge and information,
the capacity to create and spread value), in active communities
that innovate, and therefore create.
The economic models on which the publishing
industry has been based hitherto remunerated the medium less than
the content: a book by Agatha Christie costs the same as a book
by a much less well-known author. The vast majority of authors
do not live on their books. Is it not a mistake to say that we
manage rights whatever the medium? In essence, we manage a service
(as far as databases are concerned, we pay access time).
Systems of collective thought are based on a
market economy in which the transaction is king. Lawyers, legal
experts and accountants help the players to maintain their position
in this system and mindset at all costs, since it will enable
them to increase their income. Existing systems defend the rights
of intermediaries, not really those of creators and beneficiaries.
Intellectual property is the stage on which
international commercial interests will confront each other over
the next five years. If Europe wants to introduce a system which
allows creative artists and beneficiaries to appropriate the objects
of its heritage, culture and education, there is a real challenge
to be met. It is in that spirit that the Council of Europe is
working on an instrument to defend the cultural heritage against
globalisation.
The problem is a complex and difficult one.
We should not be afraid of taking a slightly Copernican view of
social transformation, in which it will not be possible to keep
on creating wealth. Creating wealth increasingly means capitalising
on the intangible and drawing on ideas of collective identity,
essential to which are ideas of heritage and the digitisation
of cultural information rapidly accessible to the greatest number.
In classic economic theory, if the marginal
cost is zero there is no price. Should we construct the rules
of the new economy on those of the old - as we are desperately
seeking to do - as though a creative work was like a car? Moreover,
the law develops according to social consensus. An economy cannot
be built on keys if no-one wants to use them. The actors in the
value chain need to be clearly identified, as do the sources of
income for each one.
The importance of a comprehensive approach
The Audiovisual and Multimedia Research and
Innovation Network (RIAM) was created in early 2001 in the wake
of the PRIAM programme and is one of the Technological Research
and Innovation Networks (RRIT) . Its aim is to take an interest
in content and usage, with the underlying idea that internet content
could be one of the drivers of high-speed access.
RIAM's objective is to bring together several
communities of manufacturers around flagship players like Thomson,
Thales, Canal Plus Technologies, etc. The network also includes
the creative aspect (audiovisual, publishers' multimedia divisions)
and research laboratories working in the sphere of digitisation
and audiovisual or audio content.
RIAM's two chief areas of research are:
- the protection of intellectual property rights and digital rights
management (DRM);
- self-production (the spread of mobile telephony and SMS has
shown that network growth is due in large part to self-produced
content and inter-personal exchanges).
The internet is not only a vector for piracy
but can also help:
- to promote better management and use of rights, through new
usages that existing media do not allow (certain sites sell pieces
of music or videos with forms of rights management adapted to
different types of use: for one-time use, one week, etc.);
- to apply the pay-TV or pay-per-view system to the internet,
via decoders, smart cards or the recent Smart Right system (developed
by a Thomson-led consortium and supported by RIAM) with the aim
of tightly controlling rights management, including on domestic
networks, ie, for all uses in the home.
The first problem with DRM is security. Beneficiaries
are of course careful to ensure that the system cannot be pirated
too rapidly (for example, CSS coding for DVD was soon circulating
on the internet). Smart cards are a functional solution for online
transactions. The protection of privacy and data is another issue,
as is the question of open-source software, bearing in mind that
one of the first security measures in this respect consists in
not publishing the algorithms (even cryptography patents are rarely
published).
Lastly, the network's influence in the international
sphere should not be forgotten. French players must be able to
disseminate their French-language content in a sufficiently accessible
way and allow for a proper utilisation of rights so as to ensure
that they are not submerged on the internet by content in English,
or even Quebec French.
As far as artistic and cultural knowledge is
concerned, a wide-ranging and ambitious approach to the use of
new technologies must be taken. Invention must genuinely defend
creative artists (rather than intermediaries) and the collective
aspect, ie, identity and knowledge-sharing (teaching, training,
open systems, etc.). Short-termism often seems to be the order
of the day, with systematically defensive positions where rights
are concerned. It is crucial to review the fundamentals of a new
economy based on paperless works and documents, for which no definitive
elements exist to date.
Possible solutions
The question may be approached from the following
standpoints:
- better prevention;
- consumer education;
- effective procedures.
- As far as better prevention is concerned, wouldn't better encryption
be the right solution, with the help or under the aegis of a public
organisation? Public information and control by a supervisory
authority seem important. For education, the question of free
information for educational purposes needs to be analysed in detail.
- For procedures, it is important to encourage internet access
providers to step up their prevention and enforcement measures
with regard to the content they host. The telecommunications crisis
is very beneficial here, since the number of IAPs will dwindle
and it will hence be much easier to require them to control the
messages they circulate.
- Lastly, Microsoft has not greatly suffered from content piracy.
Constant innovation should be a priority, since it can often be
a workable substitute for rights management.
It is instructive to compare what Microsoft
and MIT are doing. MIT has put its courses online via open crosswave.
Its revenue model is different from that of Microsoft, which is
able to achieve 85% margins despite being one of the most extensively
pirated companies in the world (its business model takes piracy
costs on board). Is protection at all costs a solution? Doesn't
a model exist with a different remuneration system which would
benefit consumers and not just the majors?
ROUND TABLE n° 3
VOICE AND IMAGE ON IP: TECHNOLOGICAL CONVERGENCE
Introduction
Convergence is one of the myths that has been
around almost since the birth of the internet. But if convergence
is defined as the possibility of providing voice/data/image services
via the same medium, it is plain that for the most part voice
networks carry voice traffic, television networks carry video
and data networks carry data. The question this round table has
to address is therefore whether for economic and technical reasons
the horizon will continue to recede or whether convergence is
becoming a reality, driven by the emergence of a single technology
in the form of the IP protocol.
Voice on IP
A number of factors need to be taken into account
in order to clarify this issue.
- Standards now exist, allowing for voice transmission on IP
under acceptable conditions, and such services have been rolled
out in certain cases.
- High-speed access is spreading, thanks to cable networks and
ADSL.
- Cable operators have introduced a number of standards which
make it possible to envisage voice transmission via cable, alongside
existing data transmission facilities.
- A number of economic factors may also help to drive nascent
competition in voice/data/image transmission.
Among telecom operators, competition between fixed and mobile
services is causing slower growth in wireline telephony and may
encourage operators to extend their offerings in terms of both
services and video content.
From an economic standpoint, convergence would enable operators
to cut their costs, by using the same medium to provide two services.
- Competition, always fiercer between cable and digital TV, may
lead cable operators to diversify their services, as they already
have done with high-speed access. Telecom and cable operators
are in competition over high-speed access. The question is whether
this will spill over into head-on competition for voice services.
- However, telecom operators could respond to this encroachment
onto their market by offering video. The range of possible dynamics
where competition is concerned suggests that convergence in this
direction is emerging, with interpenetration between services
and networks.
Newcomers onto the market will even be able to offer voice services
on already highly competitive broadband networks.
- Voice on IP already exists. If it were to be available on a
larger scale, the question arises whether the service will be
of at least equal quality to that of conventional wireline networks.
- Although a number of standards exist, voice on IP mainly uses
proprietary systems, which explains why it has been taken up first
in corporate networks and for the interconnection of such networks.
It is marginal in terms of general consumer access, even if a
number of technologies can carry voice traffic on broadband.
- As far as new services are concerned, widely-used mobile services
(messaging, enhanced directories, teleconferencing, etc.) could
be offered on fixed networks using voice on IP.
Video on IP
The technology for video on IP is now mature.
Video compression technologies (MPEG-2, MPEG-4) and video transmission
speeds are becoming compatible with the access technologies now
available. Speeds are now reaching 2 Mb/s, giving quality comparable
with that of broadcast television.
The potential penetration of high-speed networks is very great.
Television on conventional telecom networks paves the way for
services that have made relatively little headway on cable, such
as video on demand and all other peer-to-peer interactive services.
The opportunities of television could enable telecom operators
to make up for some of the slower growth in conventional telephone
services.
However, some factors of uncertainty remain
as to the spread of these technologies :
- At least 6 Mb per household : The first question is whether
access networks will permit services comparable with those that
currently exist on cable and satellite. Will it be possible to
transmit at least two television programmes on the same high-speed
channel, which presupposes 6 Mb per household, counting the related
services? The services on offer in France have not yet attained
that level.
- A wait-and-see attitude : A number of video compression
standards allow for the migration of video to telecom networks.
However, these standards are evolving and technology shifts may
be expected, with compression algorithms that exceed current capacities.
In a context of constantly advancing technology, the players concerned
may prefer to wait and see what happens.
- Content security : Content providers are reluctant to
see their content circulated on these networks, knowing that illegal
use could attain the scale of Napster in the music world.
Lastly, the question arises whether the overall business model
of these new TV distribution methods allows for prices that are
competitive with those of current providers (cable and satellite),
taking into account:
- the price of content;
- the price of networks;
- what consumers will have to spend on terminal equipment in their
home.
Experiments suggest that a balance can be achieved.
One of the underlying questions which the participants in the
round table should address is whether the economic and technological
conditions described above will enable networks using the IP protocol
to become the primary universal channel for carrying voice, image
and data.
Will the economic benefit of migrating all services to the same
network, or at least the same technology, be on the same scale
as the necessary investment? In other words, is there any real
economic advantage in convergence? Doesn't the use of dedicated
networks represent an economic optimum? Does a regulatory framework
exist within which voice can migrate to cable and television to
telecom networks?
Discussion
Convergence is a technological proposition.
But where technology proposes, the market and usage dispose. There
is a widening gap between what technology is capable of and how
it is used. What conclusions can be drawn, and how can the situation
be remedied?
The question is whether unified networks are even appropriate
for multimedia applications. Means of communication work not by
substitution but by addition. Technological convergence will perhaps
result from rather different types of use than those found on
today's networks (which will doubtless be around for a long time
yet).
Technologies
Background
IP is designed to federate, for fixed as for
mobile. However, the infrastructure needs to be adapted to multimedia.
In technological terms multimedia, especially synchronous, is
at the opposite end of the spectrum to IP: it is by nature contrary
to best efforts, to intelligence at the extremities of the network,
to equal treatment of applications and even to the peer-to-peer
principle. IP is currently adapting to multimedia. This adaptation
is taking place at different levels.
- The pioneers of synchronous multimedia on IP took an end-to-end
approach to adaptation. In other words, they adapted the speed
of the server to that of the client; that was the basis for present-day
audio and video streaming. The question of increasing the speed
is still a live issue.
- The next step was to refute the egalitarian approach, ie, to
treat different flows in different ways. Initially, each flow
was treated differently, with the result that complexity in the
network increased needlessly. The problem was partly solved by
introducing classes of services, with different priorities and
characteristics.
- Multicasting solved the problem of peer-to-peer. CDN architectures
played an important part in bringing audiovisual content closer
to clients. Servers were placed in the network to relay audiovisual
content and bring it closer to clients.
Another solution consists in using a medium other than IP. An
ATM part still exists in telecom networks to handle this type
of configuration.
xDSL technologies have an adaptability that is particularly well-suited
to higher speeds and symmetrisation.
The future of IPv6 is also important for multimedia on IP. The
main issue remains that of quality differences according to the
service (games, videoconferencing, etc.). Considerable progress
is still needed in R&D, standardisation, operational implementation
and operability.
Technological performance, especially speed, is still a real
issue.
Consumer maturity with regard to services on IP is a very important
factor, and maturity is increasing. A user playing Everquest on-line
on Sony's dedicated site can see a bar in the top left corner
of the screen which is green, yellow or red according to the state
of the network. The bar indicates:
- the time lag between you and the game server,
- packet loss in and out between you and the other player.
Players now have a high degree of technical maturity with regard
to IP transmission and may change operator during the game according
to transmission quality. This raises the question of the quality
of low-cost networks.
Does it cost much more to deliver 200 Mb than 2 Mb? The price
per bit transported has tended to increase recently. Is this consistent
with much higher speeds all round?
There is no country where operators do not hold back part of
the ADSL bandwidth so as to be able to offer it the following
year, bearing in mind that the price of low band (which carries
voice and possibly also low-speed data) is steadily falling. For
TV, it is in the interest of satellite operators to offer television,
which from the outset requires 2 Mb, if not 3 or 4. The speed
is not the same, but that should not concern the internet.
Usage
Although the technology carrying voice, data and video is converging
in terms of infrastructure and technical solutions (involving
across-the-board digitisation of content, protocols, infrastructure,
etc.), business models and usages for voice, data and video are
still different. This fact is under-estimated, especially by players
in the telecommunications industry.
The difference in business models is due to the fact that audiovisual
and multimedia services are not like staple goods that you buy
in a supermarket. They are packaged and offered to consumers through
the filter of content, which is by nature editorial. There is
an editorial approach which looks to assemble a certain amount
of raw material in order to make a product that has its own identity.
Success is the reward for intelligent packaging, which offers
something extra, of an editorial nature.
Voice on IP
The technology needed to provide voice services
on IP is mature. Compression is efficient and the bandwidth on
copper is available. There is no real technological problem.
Voice on IP is possible in the broadband data channel, whether
ADSL, wireless LAN or cable. With voice on IP, a single copper
line or cable can carry several simultaneous calls. In other words,
one household can conduct five calls at the same time, each with
its own number, service and filter. It is also possible to offer
exactly the same services as are available on a mobile (SMS, voice
messaging, etc.). Small terminals are appearing, called NOS-IP,
which can be placed in a child's bedroom, for example. They can
display SMS and it is possible to buy minutes or unlimited time
for the local zone. Voice on IP is thus giving fixed telephony
a new lease of life. However, the immaturity of operating standards
is an obstacle to the complete replacement of all consumer voice
telephone technology by voice on IP.
Audiovisual content
Television offerings are driven by three main
lead products: film, sport and news.
Films are the prime lead product. From a convergence standpoint,
the solution would seem to be video on demand or pay-per-view
- the possibility of watching any film at any time. Yet from the
standpoint of both the consumer and the content provider, the
reference in this field is the DVD, well ahead of broadcasting.
DVD now generates content producers' biggest revenue stream:
doubling every year, it drives the industry. A DVD can be rented
for €1 or €1.50 and, if it is well-made, projected onto
a 4m x 4m screen with close to cinema quality. Consumers generally
perceive only part of this quality. One euro for an hour and a
half of superb quality digital image and sound is an economic
benchmark which confronts players in the telecom sector with the
following question: when will it be possible to retail the equivalent
of 4 Gb for one euro on a mass market?
Sport, the second driver, can be taken together with news. Sport's
value lies in immediacy. Not many people watch a two- or three-month
old football match. The corresponding business model therefore
represents an attractive opportunity for players in the telecom
sector. However, a way of packaging the product attractively still
needs to be invented. The immediacy value of news is even more
obvious. There are more innovations in this sphere, or at least
more investment by specialist producers. But although someone
might watch a news flash for five minutes on a mobile phone, it
is harder to see them doing it for a whole match or a film, simply
as a matter of comfort or screen size.
Without these drivers it is difficult to build a durable economic
model for audiovisual services. The overall offer consists mainly
of lead products - which are the raw materials - onto which packaging
and the editorial line are subsequently grafted.
Radio, as a popular medium, has not suffered in the slightest
from the enthusiasm for downloading music from the internet. With
a plethora of stations, radio has successfully preserved a certain
complicity with the audience, a message and an editorial direction.
In the video sphere, cable operators have come to terms with
the aggregation of content better than telecom operators, probably
for cultural reasons. If the telecom sector really wants to get
to grips with convergence and develop offerings that meet consumers'
expectations, it will have to achieve a better understanding of
the reasons behind aggregation and the consumption of audiovisual
content.
Home use
Uses in the home include not only television
but also photography, music and video.
-The explosion of digital photography has automatically
triggered exchanges on the network, both between individuals and
in order to access services such as printing. It is currently
one of the most important factors driving the growth of services
on IP.
- As far as music is concerned, the number of portable digital
players in the United States is expected to rise to 15 million
in 2003. Download sites are becoming increasingly user-friendly
and are likely to expand.
- Hard disk capacity is growing much faster than Moore's law.
By 2008, the average hard disk will contain a terabyte of data,
equivalent to one month of continuous MPEG-2 video. Hard disks
will be found in video recorders, game consoles, portable digital
players, etc.
- Multiple PC ownership is very important for the home use market.
300,000 French households currently own more than one PC. Home
networks will soon exist, raising the question of outside connections.
Wireless technology will certainly play an essential role in exchanges
within home networks.
- Game consoles that can be connected to the network have recently
appeared (X-Box, Playstation). Consumers on this market are particularly
demanding in terms of network speed and service quality.
All these devices will have to exchange with
the outside world, which will require higher speeds. Gateways
will appear, linking devices to the outside. Their functions will
be :
- connection to the network,
- security, especially with firewalls,
- address translation (while waiting for IPv6),
- management of services and the different levels of service quality
required.
Business multimedia applications
Businesses use very few multimedia, voice and
animated image applications at present.
- Enhanced peer-to-peer communication (a form of chat adapted
to the business and enhanced with text, sound and image) is expected
to become much more widely used in the future.
- Top-down communication, which today still often relies on conventional
audiovisual networks, is another factor.
- Videoconferencing and document sharing will expand rapidly.
- Visits to pages incorporating video will also expand, constituting
a typical e-learning tool.
Through the RNRT (National Telecommunications
Research Network), a number of these applications have already
been tested. As quality improves, videoconferencing is being taken
up more widely, especially for distance learning and teamwork.
France Telecom has recently tested what it calls a "telepresence
wall", a life-like remote communication facility with a spatial
audio system that allows for several simultaneous conversations.
What stage have these usages reached?
As far as video on DSL is concerned, 2002 saw
more trials than real roll-outs. The number of global trials has
increased considerably, from five to around forty.
Some operators are highly ambitious:
- Yahoo Softbank and Tellus are each aiming for 600,000 lines
next year,
- Belgacom is also aiming for a large number of lines.
In most European countries, at least one operator (TV or telecom)
is aiming for several hundred thousand lines within whatever timeline
their respective regulator allows. The most advanced countries,
like Japan, the United States and Canada, have been the first
to pass the hundred thousand lines mark.
Cable operators have been rolling out voice on IP in a big way
- which is hardly surprising given that for them it is the lowest
cost option. Voice on IP has started on ADSL and is spreading
fast; launched a year ago in the business market, it is now being
taken up by young people in the home
Recommendations
Broadband gap
South Korean schoolchildren start handing in
their schoolwork via PC, with animated and musical illustrations,
from the age of about eight or nine. The broadband gap has been
closed very quickly in South Korea: all schoolchildren need broadband
at home in order to do their schoolwork.
In France, the gap appears at the age of 13 or 14, when young
people start wanting to download music, games or video. Everyone
is responsible, in social and geographical terms, for closing
the broadband gap, which goes well beyond the digital divide.
Closing the gap is also in the economic interests of the nation,
insofar as the broadband penetration rate and the increase in
traffic per subscriber (or revenue per subscriber) are drivers
not only for operators and manufacturers but also for content
providers (schoolwork, small business content, efficient teleworking,
access to video content, etc.).
A business model for this content exists. Of course, the number
of operators who can share the cake in a deeply rural area is
open to debate; overall, however, given perequation between urban
and less densely populated areas, the business case is
solid and drives an economy that goes well beyond just telecoms.
It is to everyone's advantage to promote this emerging market.
Regulatory disincentives
A regulatory problem arises in relation to broadband
video: that of the division of rights between telecom operators
and audiovisual producers. Some confusion reigns over the scope
and powers of the audiovisual and telecommunications regulators
(CSA and ART). A certain degree of maturity is needed so that
a distinction can be drawn between commercial content or programme
bundles and peer-to-peer multimedia communication. Legal grey
areas still exist, such as the obligation on mobile operators
to open their networks to virtual operators or France Telecom's
right, as a carrier, to offer television on its networks.
ROUND TABLE n° 4
BRINGING EUROPE OUT OF THE TELECOMS CRISIS
Introduction
- The telecoms crisis has its origins in four inflationary bubbles
:
- Too many operators over-invested in identical markets. Mainly
CLECs in the United States and alternative long-distance operators
in Europe, they committed too much capital. The investment curb,
linear until the last two years, suddenly accelerated.
- American pension funds invested heavily in new technologies.
Investments were frequently redundant, with several firms being
given money for the same type of project, and did not follow a
rational overall strategy.
- European mobile phone operators paid astronomic sums for UMTS
licences, even though the uses of high-speed mobile telephony
have not been clearly identified.
- Operators and manufacturers embarked on a takeover spree just
when stock prices were at their peak. Most operators sought international
expansion, acquiring foreign operators at high prices which they
had to write down the following year.
These four bubbles, each of which represented 200 to 300 billion
dollars, all burst at the same time. It is now acknowledged that
investment cannot pick up again of its own accord. The effects
of the bursting of the capital investment bubble will be absorbed
in time, but companies are under-investing because they lack financial
resources, even though needs have been clearly identified.
- Despite the crisis, customers have not stopped using telephones:
traffic, both fixed and mobile, has actually increased. The recovery
depends above all on rising demand in new areas. Strong growth
in broadband use should create a European high-speed market worth
140 billion euros by 2006.
- 40 billion euros from the connectivity market (operators receiving
monthly payments for ADSL, wireless LAN, etc.);
- 40 billion euros from content (TV, e-commerce, etc.);
- 60 billion euros from indirect growth (content production generates
needs for software houses, hardware, people, etc.).
- In the United States, a huge public fund has
been earmarked for reviving the telecom sector, targeting two
segments:
- research,
- manufacturers, then operators.
No similar decision has been taken in Europe, although the industry
is lobbying hard.
- The first objective of any initiative must
therefore be to encourage usage.
Governments in many countries have fastened
on to this objective, committing themselves financially to the
rapid spread of information and communication technologies.
- In South Korea, the internet is widely used in public education.
A child without broadband in the home cannot learn English and
cannot therefore get into a good elementary school. All South
Korean parents have ADSL. It is probably not possible to transpose
this kind of approach into France immediately. However, the South
Korean government's investment in education has been enough to
lift the corresponding industry into orbit.
- The United States sets great store by e-government, meaning
the possibility of conducting all sorts of government business
(social security, tax, applications for hunting and fishing licences,
etc.) online. A number of e-government initiatives require broadband.
Other applications being developed encourage
people to get high-speed access :
- tele-surveillance: many cameras linked to centres by broadband
have been installed, especially as anti-terrorist measures. At
an individual level, this type of system is widely used for elderly
people;
- tele-medicine which, already advanced in ISDN, really only works
properly with broadband.
The fundamental question is what percentage
of subscribers should broadband reach, whatever the technology.
A target of two-thirds of the population (the official US government
target, which South Korea has already achieved and which Japan
will have reached by the end of 2003) implies a year-on-year penetration
rate of 7%, which is not unreasonable. However, all the players
concerned (government, operators and manufacturers) need to share
a common objective.
- If the spread of new technologies is to be encouraged, the factors
holding back network use and the development of content and services
need to be identified. Lack of confidence among content providers
severely restricts innovation. Although confidence exists in areas
such as teleworking, distance learning and e-government, that
is not the case for televisual content. The transfer of rights
to telecom operators has not been perfectly negotiated and will
probably require arbitration by the government or public institutions.
- In research, efforts need to be made both at European level
and, more specifically, in France. France is not among the top
20 developed countries in terms of broadband, internet or ICT
use. It is aberrant not to take steps to remedy the situation,
especially as French industry is well-placed and France is among
the world's top six nations in many other areas.
Discussion
Introduction : Discussion of the causes
In the US in the 1970s, a yawning gap appeared
between the capacities of NASA and other defence agencies and
a very dated telecommunications network. The potential of communication
technologies existed but was not being exploited. The whole aim
of deregulating the telecommunications market in the US was to
create forces that would necessarily lead to introduction of the
most modern technologies.
In Europe, in contrast, telephone use was rationed. Seven or
eight years ago, an American telephone generated five times more
traffic than a French telephone. The same technologies are used
on both sides of the Atlantic but with very different business
models, causing Europeans to limit their telephone consumption
and use. Telephone use in the US expanded rapidly in the 1920s
and 1930s, whereas it did not really take off in Europe until
the 1970s.
Growth has been steady in the 15 years or so since the arrival
of the internet. The internet has considerably facilitated the
process of innovation through the adoption of standards which,
though rustic, are common and accessible to all. The internet
has therefore made it possible to realise both the gains arising
from technological progress and productivity gains (which are
still appearing and will continue to do so, through job losses
and the creation of new jobs). The internet will revolutionise
finance, banking and retailing, though only slowly since mindsets
do not change as quickly as technology. But it is plain that the
internet attracts free content and prospers because of it, making
it difficult to sell expensive services.
Value has moved elsewhere in recent years, to DVDs and peripheral
applications. Telecom networks are just pipes. Information, which
now represents most of the value, is a highly unstable commodity
that is difficult to evaluate. It is all the more difficult for
usage to spread in that access to content is rationed. Business
models need to be revised and adapted to take account of these
new realities. The shift has been a godsend for society and for
the downstream economy, but it has also been a rude awakening
for operators, especially those that went too far, either paying
too much for licences (Bouygues was wise enough to resist) or
making cash acquisitions at prices that were much too high.
What should be done ?
Supporting research
Research and innovation also represent an opportunity
for Europe. It is essential that the potential represented by
manufacturers should not disappear. They are in a difficult situation
following the bursting of the internet bubble, for which they
were not responsible. It would be a serious matter if France were
to lose its know-how, in optics for example.
Research in various areas needs to be continued if broadband
technologies and multimedia services and applications are to become
widely available.
In the current unfavourable environment, equipment manufacturers
are trying to maintain their R&D, especially their research,
so that they can bounce back when the economic climate improves.
The American government has poured two billion dollars into Lucent
and another two billion into Nortel.
A number of European telecom firms have initiated discussions
and sought solutions in collaborative research. However, the investments
that manufacturers are willing to make must be supported by governments
at both national and European level.
The EU has just launched its sixth framework research and development
programme. It only partially meets players' expectations and has
two major flaws:
- insufficient public funding - a mere 500 million euros, when
the National Telecommunications Research Network estimates the
amount needed to revive the sector at 2 to 3 billion euros;
- a long-term perspective - projects are due to end around 2006-2007
and would not therefore reach the market until 2007-2008.
European telecom manufacturers have got together to put forward
a joint initiative reflecting their imperatives, which they have
called CELTIC (Cooperation for a European sustained Leadership
in Telecommunications).
Supported by all the telecom majors, the initiative takes the
form of a large-scale Eureka initiative along the lines of similar
initiatives in embedded software (ITEA) and micro-electronics
(MEDEA+). The aim is to launch the new programme in mid-2003,
with the first projects starting in early 2004 and the first transfers
of services or applications taking place in 2005 or 2006, in conjunction
with the sixth FRDP.
Academic and industrial research in France is very good. However,
the sector's reactivity is much greater than the capacity to set
up programmes. We have to look to the future. There is no advantage
to be gained from devoting public research entirely to industrial
projects, even though that is essential in order to get the focus
right. Public research must generate sufficient margins to finance
research further upstream so as to pave the way for subsequent
innovation. In major research organisations, systems for guiding
research and allocating budgets are based on publication; they
take virtually no account of industrial realities and need to
be adapted if we are to make up lost ground.
Supporting use…
… By developing a clearly defined and
suitable telecoms offering
Supply is the first stumbling block. A small
firm looking for an optical fibre link cannot find one except
at a prohibitive and unprofitable price. Likewise, radio local
loop offers have completely disappeared.
The only way of reviving the telecom sector, and to a certain
extent the IT sector with which it has relatively close links,
is to ensure that supply meets user demand. Even then, users must
not be disappointed with the services on offer and must be satisfied
with their technical quality. Operators have worked hard to promote
ADSL to small business users.
If businesses are to be convinced, however, the service has to
be flawless. There is no point in having a high-speed link between
the subscriber and the DSLAM if the operators upstream have low-speed
networks.
In order to make up for insufficient investment, the operators
need to be sure of a stable business environment. The business
case of a telecom operator (mobile, broadband or corporate high-speed
access) is viable only with a return on investment in less than
three years. Stability for at least three years is therefore essential,
for both incumbents and alternative operators.
... Through education and training…
In most countries, and more particularly in
the UK and the US, when exchanges were slimmed down most of the
technical staff became surplus to requirements. In France Telecom,
these employees were transferred to the sales side. Their particular
status meant that they had to be found other jobs in the company
(there is talk now of 45,000 staff being given early retirement
or made redundant). In a similar situation, British Telecom and
the American operators let their people go. Many of them were
recruited by SMEs, thus spreading their technical skills and contributing
to very strong demand for telecom applications.
SMEs are short of trained staff, especially network and computer
technicians. Most training takes place on the job, by retraining
employees with initial training in other fields. University technology
departments do not turn out enough qualified people. Vocational
short degree courses are excellent, but produce only a trickle
of graduates. A post-graduate course produces only ten to fifteen
people a year. There is a real shortage of managers with dual
skills (technical and professional) in firms of all sizes, since
the French education system barely addresses the need.
Lastly, public education in France is free. Educational institutions
generally have limited resources, whereas disciplines like art
history have substantial equipment needs. Fortunately manufacturers
are relatively generous when it comes to donating hardware.
... By cutting prices to consumers
Consumer demand is not restricted to the managerial
classes only. But the average French worker earns little more
than the minimum wage. So there is no hope of selling services
that cost hundreds of euros a year; instead, providers should
be thinking in terms of an amount in double figures for a bundle
of services including television, mobile phone, fixed phone and
internet. In some European countries it is possible to have a
2 Mb link for a handful of euros a month. The question is whether
French operators have the means to offer such services. The cost
of the technology decreases with mass production; in South Korea,
for example, the critical levels needed for mass production have
already been reached. In France, price regulation has caused operators
to propose only relatively expensive interconnection services,
whereas the bulk of the market is at the lower end. Only content
will revive the market: content emerges very quickly once the
pipes exist, and the population, across the entire spectrum, takes
it up immediately.
… By taking advantage of extensive mobile
phone use in Europe
Another factor in developing broadband applications
is synergy between fixed and mobile services. UMTS is late. Recent
initiatives taken by the government and the president seem to
be moving in the right direction: at European level, operators
need to be released from licence constraints, leaving the market
with a little bit more room for revival.
GPRS and broadband internet content is very similar. Content
providers also need broadband access to offer their content on-line,
meaning ADSL or Gigabit Ethernet. People using this speed are
satisfied with it.
Some initiatives in the services field, such as those linked
to Bouygues' i-mode, are likely to accustom people to consuming
more medium-band services, which will in turn encourage them to
consume more broadband services.
Should the market be allowed to purge itself
?
The collective strategic error was to think
that lasting rents were possible. On the contrary, such rents
are disappearing. The technical and economic environment is no
longer conducive to the constitution of rents except for those
of a copyright nature, like Microsoft's.
Some people argue that there is no point trying to save the historic
assets of the telecommunications industry. Operators have gone
out of business because they got things wrong or made poor choices
(WorldCom and others). But we should understand, like the United
States, that the current situation is a godsend.
Since the "cannibalistic" destruction of rents (to
use the term coined by Andy Grove, former head of Intel), the
question has arisen that Bill Gates regularly asked five or six
years ago: what happens when everything is free? Twenty years
ago, the United States wanted to move from a communication model
to a free model, from which the American people were supposed
to benefit. The American government regarded the fact that it
entailed the death of a certain number of operators as a good
thing for moving the industry forward.
The Financial Times has estimated that over-investment
in the telecom sector amounts to two to three thousand billion
dollars, impacting the accounts of both banks and operators. An
article in the newspaper expresses the hope that the European
Commission will wean the French government away from the belief
that it can save France Telecom merely by injecting cash. The
FT article argues that operators should go to the wall if necessary
in order to make the industry more efficient.
Conclusion : What market ?
What will the market be like once the four bubbles
have finished deflating? Will it be the same as before, with relatively
little competition and reasonable margins for vertically integrated
equipment manufacturers and operators? Or will there be much more
competition, low margins and a horizontal structure, with players
shifting their value added to specific segments rather than trying
to cover the whole market, as in the IT sector. The answer is
important, especially in order to calibrate R&D.
The consensus view is that the post-bubble market is likely to
be slightly smaller. Even with 3G and ADSL, few people expect
the growth rates seen with GSM three years ago. Most industry
watchers expect them to remain at the same level as five or six
years ago.
Regarding the shape of the market and the number of players,
financial analysts agree that the arrival of start-ups and new
small operators has had less effect on the market, as far as the
manufacturers are concerned, than the arrival of Asian competitors
(margins are slim and will remain so). Operators' margins will
also remain low. Yet there will be fewer players than when the
bubble was at its height. The current process of globalisation
will continue, among both operators and manufacturers, though
new entrants are expected both in the telecoms industry (including
Microsoft) and in the content industry. Some players will doubtless
leave the field.
ROUND TABLE n° 5
CRITICAL TECHNOLOGIES : PERCEPTION AND REALITY
When Gérard Roucairol and Jean-Claude
Merlin wrote their report on the internet of the future for the
National Telecommunications Research Network (RNRT) in 1999, the
model on which it was based described digital telecommunications
in terms of three worlds:
- the real world, made up of individuals, organisations and everyday
things;
- the world of communication and connectivity;
- the world of intermediaries, the necessary link between all
communication, and of on-line services, made up of software, computers
and databases.
These three worlds were distinguished by a standard, IP, stronger
now than ever, which breaks the one-to-one relationship between
the telephone subscriber, the telecommunications operator and
the service provider, making the three worlds entirely interdependent.
In addition, as the world of communication is between two standards,
so-called commodity services have become low-margin, high-volume
items.
The bursting of the internet bubble has not fundamentally changed
this model :
- In the world of intermediaries, a whole host of software, machines
and databases are interconnected on the web. The major phenomenon
is the integration of all these elements through standardisation,
through links between applications (web services) and data integration
as a result of the movement which began within W3C (web semantics).
In the internet world, where time is measured in microseconds
or nanoseconds, there is no longer any place for human intervention.
The system has to work almost autonomously.
Software infrastructure is converging, as can be seen from corporate
information systems (the market of choice for traditional IT)
or remotely accessible value added telecommunication services.
Both ERP applications and on-line telecommunication services are
accessible on the web. The two major standards are those of Microsoft
(.net) and J2EE (backed by Sun and much more open).
- In the world of connectivity, the important factor is that
IPv6 is at last beginning to spread. The widely held idea that
the core of the network is relatively stupid in relation to old-structure
networks and that intelligence is located at the extremities is
highly debatable. Intelligence is still located at the extremities,
true, but it is getting much closer to the network. In fringe
networks, infrastructure that takes into account what is transferred
by the network, the content or the applications is spreading.
- The biggest impact on the real world has been made by wireless
technology and high-speed access via xDSL or Gigabit Ethernet.
The question is whether UMTS will annex the relationship between
the individual, the operator and the service provider.
The real world is changing. The European Commission uses the
fashionable term "ambient computing" (things and individuals
become communicating objects). Embedded software is needed; the
question that then arises is which operating system and which
processors to use. Of course, if the ambient approach is to be
realistic, there will have to be sensors and actuators in all
the objects concerned. There is a definite trend in this respect
towards proactive computing, ie, an increasingly autonomous digital
world, providing assistance to the individual and with generalised
peer-to-peer relations.
Discussion
For operators, the time horizon is one of the
decisive elements of critical technologies. Some technologies
are critical now, some will be critical in three or five years
time, others may be critical in ten years or so.
Critical technologies are those that will be successful. Finding
a market implies, apart from competitive pricing, identifying
players capable of getting together in a value chain and customers
who want to use the technology, ie, who see in it the answer to
some of their deep-rooted frustrations. In this respect, critical
technology has a sociological and ethnograpic dimension.
In order to find the right business models for value chains,
it is important to test the corresponding technologies relatively
early, via technological application platforms, and to work simultaneously
on the business models.
Generic digital technologies
There are four generic digital technologies
in ICT.
- Software
Whether in mainframe architecture, infrastructure software, data
fusion or network security, a number of generic technologies are
required, software first among them. For systems or equipment
manufacturers, software represents 80% or more of the development
cost.
- Microelectronics
Certain aspects of microelectronics are in the commercial domain
(DRAM, etc.), others are much more sophisticated, especially in
very high-speed or very high-performance systems. Some highly
specific designs and applications can pave the way for leadership
technologies. Nanotechnologies are an extension of microelectronics,
though not enough consideration has always been given to future
advances in the commercial realm.
- Optoelectronics
Development of optical networks, hyperfrequency optoelectronic
interactions and their civil and military applications.
- Signal processing
Of course, these generic technologies have their own lives, but
the extent to which they are relevant or critical or leading depends
to a great extent on the progress made in each one. Sometimes
progress in one generic technology can cancel out progress in
other technologies.
For example, systems on a chip are a very important development,
combining several digital technologies (microelectronics, signal
processing, software). They represent substantial value added
and may become a decisive factor in international technological
competition.
Critical telecom technologies in 3 to 5 years
These primarily concern access technologies
and technologies giving higher access speeds.
Optimisation of speed on paired copper cable
ADSL was still a critical technology not long
ago, but the key issue now is to increase the power of ADSL networks
and to design and engineer them for millions of users. The engineering
of ADSL access networks (backbone and support) has in turn become
critical.
Various technologies are currently being implemented:
- VDSL (which allows for a significant increase in speed),
- ADSL Plus.
Each of these technologies raises specific issues relating to
the business model and value chain. Security-related aspects will
also determine whether they develop more or less rapidly.
Mobile access
Radio and signal processing technologies are
also critical for mobile access. Customers will demand greater
mobility but radio resources will always be relatively scarce.
The capacity to optimise and get the most out of the radio spectrum
will be an important factor for a long time to come.
The critical factor for 3G - leaving aside the cost of licences
- is interoperability. It took several years to achieve effective
interoperability with GSM, and it is likely to take just as long
for UMTS.
Wi-Fi offers really interesting possibilities for local networks;
the technology is also being developed by pirates (especially
in Paris).
The business models of these two technologies are rather different,
since Wi-Fi is practically free and UMTS is relatively expensive.
To what extent are they complementary and to what extent are they
in competition?
In the countries which use both, namely Japan and South Korea,
each has a substantial share of the market, especially as the
terminals are different. 3G terminals (small terminals for young
people) target a market interested in games, photography and video.
Wi-Fi terminals target professionals (PDA, PC).
Wi-Fi is entirely complementary to 3G. It offers high-speed access
in a certain number of places, without any major restrictions
on capacity (for the time being). These restrictions, like those
relating to security, need to be dealt with very quickly. With
optimisation of the spectrum, 3G should bring considerable capacity
to wider geographical areas. So far, it is entirely possible to
envisage Wi-Fi and 3G working in tandem, and relieving different
kinds of frustrations .
Access terminals
In the home, the need for interoperability between
different items of hardware and the arrival of domestic gateways
(television, set-top box, microcomputers, ADSL modems, etc.) will
increase the capacity for communication.
Increasingly, operators will act as intermediaries. Critical
technologies in this area are linking, intermediation and web
service technologies which will make it possible to assemble quickly
- and if possible openly - services from all over the world so
as to create new portals or new and increasingly complex software
suites, simply by assembling them and calling up services.
Microsoft says it believes in the continuity of service between
Wi-Fi spaces and mainly GPRS spaces, with the PDA as the terminal
of choice. A set of technologies depends on identification in
a mobile world. Information technologies are therefore probably
critical. Clearly identification, and hence access control, are
areas in which France, with its tradition of high-level mathematicians,
can stake out a position. The problem also arises of reconciling
the need for identification with respect for privacy. It is an
interesting technological challenge.
Upgrading network cores
Three elements could become critical in network
cores :
- the move towards optical technologies, delayed
somewhat recently, but likely to continue, starting with optical
cross-connecting (multiplexing and, probably one day, routing).
The key to this development is designing network architectures
from scratch, on a purely optical basis. Network engineering will
become critical. More generally, as network access and internal
network technologies change in general, network engineering or
re-engineering will become a source of substantial savings;
- peer-to-peer, important in terms of
use, technology and implementation for operators, and the associated
security issues relating to access (identification, authentication)
and data transmission;
- IP itself (it will be necessary in the next
few years to implement the multiplication of addresses on IPv6
and find the right paths for migration, which will be a major
challenge in itself).
IT
When the IT industry in the 1960s wanted new
applications, it developed them from scratch, including:
- the database management system,
- the operating system,
- the programming language,
- the memory,
- the processes,
- the microelectronics.
The industry was entirely vertical. Compatibility
problems had to be solved by microelectronics. Specialisation
became increasingly prevalent as it became more and more difficult
to cover all these different aspects. The industry itself lived
through this very rapid change, giving it great flexibility and
mobility. Software generated more and more of the value added.
The recent pre-eminence of software
The web emerged in the late 1990s. The internet
was designed with the idea in mind that software should play an
important part. Software became essential when new network infrastructure
was rolled out, to the point where the press now considers Microsoft
and Nokia to be rivals (especially over the software environment
for mobile phones). Software issues and telecommunications issues
are merging.
Players are too often reluctant to talk about the software industry.
To get a proper perspective, it is necessary to understand where
the opportunities and the value lie. In fact, the value really
lies in the software, which fulfils essential functions in the
production not only of new products and services but also of telecommunications
infrastructure, corporate information systems and the deployment
of the internet in the broad sense.
Simultaneous development and deployment
The web arrived on millions of user PCs so rapidly
that the same teams were responsible more or less simultaneously
for design and standardisation. The periodicity of production
cycles is not the same in the software industry as elsewhere.
Even as a standard emerges, product developers have already integrated
it into their concerns.
Europe's technological competitiveness
Technologies may be classified according to
the following typology :
- commercial technologies, which are renewed
and have their own lifecycle based solely on commercial criteria
(some aspects of microelectronics fit this definition);
- leadership technologies, which give a definite competitive advantage
and enable players always to be in the running;
- sovereign technologies, which are essential to the notion of
independence.
Value added is to be found more in the last two categories than
in the first.
Closing the R&D gap
A relatively important change has taken place
in French research, with public operators reducing their research
effort. Where should the initiative for new technologies come
from if not from France Telecom R&D? The big manufacturers,
start-ups, universities?
France Telecom has changed its strategy. The CNET (National Centre
for Telecommunications Studies) no longer has the mission of spearheading
the development of French industry. All the players, in networks,
assume their own responsibilities and work together on the critical
technologies of the future. Together, they also identify new markets
and new applications so that French people can take ownership
of new technologies by using them and consuming them as early
as possible. All too often, however, as can be seen from the internet
and mobile phones, consumers are not in the vanguard when it comes
to using new services. It is important to understand why.
The civil/military couple can be perceived in different ways.
There may be a movement from one to the other after the fact:
military technologies can be adapted for civil use and vice versa.
In the current context of globalisation and competition, dual-use
technologies may be developed deliberately. After essential technologies
have been identified to a horizon of five years or more, both
civil and military researchers try and find a common set of specifications
so that work on developing the technologies in question can begin.
Differentiation comes later.
Ensuring a legal framework that favours innovation
Regulation is not neutral where radio technologies
are concerned, as can be seen from the impact of the UMTS licence
fee. The issue of regulation is particularly relevant to competition
between UMTS and radio LAN. Radio LANs work at the present time
because not many people use them. The frequency allocated to radio
LANs is the same as the one used by microwave ovens and industrial
applications, but when use of the technology really takes off
and it has to cope with much larger volumes, prices will rise
(dedicated frequencies will have to be found, and that costs money).
Regulation will be needed.
- In signal processing applied to radio, ideas are surfacing,
especially in the United States, about flexible spectrum management,
with differentiation based not only on space and frequency but
also on time (certain systems being able to cohabit on the same
frequency bands). The regulatory framework in Europe is taking
time to adapt. Shouldn't thought be given to the role of regulation
in innovation and the development of technologies?
- With the growing demand for mobility, optimising the management
of the radio spectrum will become increasingly important. The
issue needs to be addressed, whatever the state of the regulations.
As far as critical technologies are concerned, the current state
of regulation needs to some extent to be disregarded. The technologies
and the possibilities they allow often conflict with existing
regulations, which then need to be reassessed.
Patents have been widely used in the telecommunications industry;
the audiovisual industry is highly sensitive to copyright. When
a new industry is invented, it is by no means certain that the
legislative and regulatory environment appropriate to sustain
it will be exactly the same as the one developed in previous years
for other needs.
ROUND TABLE n° 6
CONFIDENCE AND USAGE: PERCEPTION AND REALITY
Introduction
It is generally accepted that lack of confidence,
whether on the part of consumers, users or vendors, is holding
back the development of a certain number of uses of information
and communication technologies.
- Is this problem rooted in the reality of technology and behaviour
or merely in perceptions? Is it a question of confidence in
the internet itself, in the security of exchanges and payments,
or in the good faith and ethics of the commercial and administrative
interfaces that users encounter ?
In the early days of consumer e-commerce, service quality
was so poor that customers would have shunned the providers
even if payment systems had been absolutely secure. Other negative
factors include the use of personal data, spam, etc.
- Is it a question of confidence in the confidence mechanisms
themselves?
It is an issue that concerns backdoors and labels and the capacity
of labels to enforce charters that companies have signed up
to and to become better known.
Two initial questions arise :
- the existence, in particular areas, of real
stumbling blocks, real difficulties that have to be resolved;
- among identified stumbling blocks, the possibility
of collective, public action (in terms of research, regulation,
awareness raising, etc.).Dan un premier temps, deux questions
se posent :
Discussion
Secure payment
Although security is not sufficient in itself
to ensure wider use it can contribute, especially if it is guaranteed.
If bank cards are accepted by more and more retailers, it is
because they are guaranteed payment.
The security of a system as a whole depends entirely on the
security of its weakest link. With the same users, the same
cardholders and the same card issuers, the level of security
is not always the same when payments are made from terminals
in different countries. For example, Thales makes PIN pads with
different levels of security for different countries. Visa has
worked with the American laboratory Infoward to define methods
for evaluating aspects such as the confidential code protection
level in PIN pads.
It would be necessary, at least at European level, to define
through common criteria (identified by the 15-408 standard)
protection profiles and public security methods accessible to
the competitors involved. Payment terminal users must also agree
to pay the small additional charge that guarantees security.
Encouraging bank card use could also have a significant effect
on the growth of on-line transactions. Transactions paid by
bank card currently account for only 7% of all transactions
worldwide. From a macroeconomic standpoint, embracing all the
players concerned, bank card payments are much less costly than
cheque or cash payments. As well as generating productivity
gains, initiatives to encourage bank card use would probably
also have a significant effect on the growth of on-line transactions.
With smart cards and expertise in electronic transactions,
Europe has acquired real technological expertise in security
matters, and insisting on a certain level of security may also
be tantamount to protecting an industry.
Personal data protection
According to some, the EU Electronic Signature
Directive has been implemented in certain countries in a way
that is incompatible with any possible return on investment.
The project's requirements are too stringent at the current
time in relation to what can be done to support the players
on the market. This criticism is all the more real in that in
Asia most people, or at least manufacturers, care little about
privacy issues. In the United States, the protection of privacy
is a personal rather than a government matter.
Nevertheless, because there is no broad data protection legislation
in the United States, the government has introduced the Child
Online Protection Act (COPA), a very restrictive law on internet
and young people. Fifteen states have also introduced tough
laws (much tougher than in Europe) on spam. A privacy issue
has emerged in the national consciousness and the US is beginning
to regulate things that it is not in the habit of regulating.
Biometrics is another case in point. The European Commission
has launched and funded a biometrics initiative, considering
that the technology improves privacy protection. Although the
CNIL (the French data protection agency) is concerned about
the existence of any centralised and exported database of digital
signatures, recording a digital signature on a smart card does
not seem to pose a problem. An important study of the subject
is published on the CNIL website.
In 2001, in the framework of a Proxy programme, Thales developed
a biometric match-on-card terminal (ie, with a control inside
the card) which can authenticate the holder before signature,
meaning that the card is not transmissible. The technology uses
a silicon bar sensor which does not retain the fingerprint.
The technique should satisfy the CNIL, but the technology is
much more complex to implement than technologies based on a
confidential code.
In July 2002, the CNIL launched a spam mail-box for complaints
from internet users. 350,000 e-mails were received in less than
6 months.
The link between confidence and security
The key problem with regard to security is
reparation. Security failures may cause harm to consumers or
providers. It is also possible to be responsible without being
at fault.
A new problem will arise with new systems, ambient technologies
and ambient intelligence: that of safety. The precautionary
principle, in connection with any safety problem, will probably
be one of the major issues in coming years.
The context of confidence is that of an organisation of society
that facilitates sharing. The knowledge-based society is a society
built on trust. The challenge is that of transparency, which
is of a quite different order of difficulty. Where confidence
is concerned, security and transparency are essential. Transparency
is the founding principle of data protection legislation. There
will be confidence if there is transparency and if everyone
involved is committed to it.
The issue of confidence arises essentially in relation to novelty
and the unknown. When buying by mail-order, French people use
the post office and give their bank card number over the phone
without a qualm. But they are very reluctant to use networks
and internet interfaces, never mind WAP interfaces. The lack
of confidence among users is mainly due to the fact that they
do not find the references in the virtual world that are familiar
to them from the real world.
Confusion exists with regard to internet security standards.
The general public believes that it is possible to rob a bank
via the internet. In the healthcare sector, staff send faxes
without a second thought and leave confidential documents lying
about. People send written social security forms by post. There
are no leaks. The problem of security is therefore one of consensus
and not only one of technology.
There is no such thing as zero risk. Confidence is therefore
a matter of belief, starting with belief in the capacity to
make good any harm that might be caused. Some damage can be
easily and totally repaired, for example by a reimbursement.
Other types of damage, like invasion of privacy, are less easily
reparable. Introducing mechanisms for repairing damage implies
solving the problem of moral hazard (ie, deviant behaviour by
a part of the population, such as seeking reparation without
a rightful claim).
- Confidence in systems (computer systems, telematic systems,
payment systems) requires first and foremost confidence in those
who operate them. Confidence in others means having a relationship
of trust with an institution.
- In certain sectors, better security does not necessarily
mean greater use. Better road safety has not increased road
use. The very notion of confidence or trust relies on the individual.
Studies have shown that confidence profiles are linked to usage
profiles. The heterogeneity of the population is a factor that
lessens confidence. People in the south are less trusting than
people in the north. Education levels also have an influence
on confidence levels.
The aim of confidence is growth. Confidence is in fact a marketing
term now used to judge security. One way of getting the public
to accept the importance of security and of paying to guarantee
security would be to confront them with the consequences of
a serious security failure. It may be extreme, but it would
surely be effective. Demonstration by example is the most effective
way.
Options for action and conclusions
Although web browsers are quite complicated,
the software for making online payments is relatively easy to
understand. But once an online payment has been made, the interface
becomes totally opaque and it is difficult to find someone to
contact if there is a problem. Vendors should be under an obligation
to provide a recognised contact point or interface.
Education
The first action is educational. In the decree
of 17 July 1978, the CNIL is given the task of advising "persons
and organisations", a task which it regards as particularly
important. Even before developing filtering or anti-spam tools
(not effective enough at present), a considerable amount of
awareness-raising and education is needed. The CNIL held a press
conference on the subject of spam last week, at which a module
explaining the legal and technical issues was presented.
- Consumers need reference points. The vast majority of consumers
are not necessarily able to keep up with the pace of technological
progress. A web browser is a very complex tool for most people.
- For retailers, the priority is to sell and deliver. The last
thing they want to worry about is implementing a payment solution
from among the 300-400 that currently exist. They need to be
able to guarantee that their customers will be able to use the
system.
- The job of telecommunications operators and manufacturers
is to transport data and create standards and standard products.
The industry is being asked more to agree on standards than
to create different forums, which only add to confusion, even
within the industry.
- The government should be the last to intervene,
and then only when things go wrong. For example, it would be
particularly unhelpful at present to push for the labelling
(ie, certification, in the terminology of the French consumer
affairs and fraud control authority) of online means of payment.
Terms like "control", "authority", "regulation",
"right" and "penalty", more appropriate
to a police station, crop up in much of what has been said so
far. That kind of vocabulary is incompatible with confidence.
Education should focus as much on risk as on labelling. If the
world is a less certain place than it used to, it should be
shown to be so.
Educational initiatives should therefore target the industry
and institutions, not just individuals.
- For example, all too often circulation lists continue to
show the e-mail addresses of all recipients.
- Likewise it is important to remind the professionals that
everyone in their private life has a right of opposition. Article
27 of the data protection act states that people must be informed
beforehand when information about them is collected. There is
also a right to access this information and rectify it, and
a right of opposition with regard to persons who pass that information
on to third parties. These obligations already existed in the
real world.
Regulation
A bill on the digital economy, currently being
debated, is designed to transpose the 2000 EU Electronic Commerce
Directive into French law. It will validate the double-click
principle for contracts. It will also transpose the 2002 Data
Protection Directive, and as such will deal with unsolicited
electronic mail.
Consultation
Consultation has increased between authorities
and online retailers and, more recently, between consumer groups
and operators, but often within restricted forums.
- It is important to create frames of reference for payment
solutions. Confidence in a currency is a specific thing, and
absolutely essential in regulating an economy. It is not possible
to transpose all the measures that need to be taken to ensure
confidence in the currency to mechanisms intended to ensure
confidence in other goods and services.
- Lastly, for both consumers and professionals, it is important
to come up with a form of labelling that is the fruit of the
consultation and the frames of reference described above. Labelling
and certification are certainly necessary. However, the problem
is that there may well be too much, which would not generate
confidence either. Coordination is needed in order to avoid
confusion.
In conclusion, there is no demand for new
technical or regulatory measures. The focus should be on providing
information and raising awareness. Confidence should also be
approached from a positive standpoint, meaning that it should
be regarded as something that is built up in a relationship,
and not from a purely defensive position.
|
