Ethical Implications of Emerging Technologies - Unesco 2007

7/31/2019 Ethical Implications of Emerging Technologies - Unesco 2007

1/90


2/90

Prepared by Mary Rundle and Chris Conley

Geneva Net Dialogue

UNESCO: Paris, 2007

Ethical Implications

of Emerging Technologies:

A Survey


3/90

The ideas, facts and opinions expressed in this publication are those of theauthors. They do not necessarily reflect the views of UNESCO and do notcommit the Organization.

Recommended catalogue entry:

UNESCO. Information for All Programme (IFAP). Ethical Implications of Emerging Technologies: A Survey .

Edited by the Information Society Division, Communication and InformationSector (Editor: Boyan Radoykov) Paris: UNESCO, 2007 - 92 p ; 21 cm.I - TitleII UNESCO

III Information for All Programme

Published in 2007 by the United Nations Educational, Scientific and CulturalOrganization,7, place de Fontenoy F-75352 Paris 07 SP, Paris, France

Coordinator of the publication: Boyan Radoykov

UNESCO

All rights reservedCI-2007/WS/2 - CLD 31112


4/90

ETHICAL IMPLICATIONS

OF EMERGING TECHNOLOGIES:

A SURVEY

Prepared by Mary Rundle and Chris Conley

ACKNOWLEDGEMENTS

In the course of compiling this report, a number of prominent members of theworldwide technology and infoethics communities have shared their expertiseand perspectives with us, including: Hal Abelson, Ben Adida, Ang Peng Hwa,Kader Asmal, Ted Barnett, Stefan Bechtold, Peter Berghammer, Scott Bradner,Stefan Brands, Kim Cameron, Shunling Chen, David Clark, John Clippinger,Urs Gasser, Lauren Gelman, Georg Greve, Dick Hardt, Dewayne Hendricks,Chris Hoofnagle, Hai Huang, Ben Laurie, Ronaldo Lemos, Lawrence Lessig,

Jamie Lewis, Abdel-Hamid Mamdouh, Dsire Miloevi, Nicholas Negroponte,Cory Ondrejka, David Reed, Eric Rescorla, Jonathan Robin, Judit Rius Sanjuan,Wendy Seltzer, Lara Srivastava, William Terrill, Paul Trevithick, Jake Wachman,David Weinberger, Ian Wilkes, and Jonathan Zittrain.

* Mary Rundle is a fellow at the Berkman Center for Internet and Society at Harvard Law

School and a non-resident fellow at the Center for Internet and Society at Stanford Law

School. Chris Conley holds a masters degree in computer science from M.I.T. and is currentlya J.D. candidate at Harvard Law School, where he focuses on the intersection of law and

technology. This paper has been produced under Geneva Net Dialogue, a Geneva-based,

open, international association whose mission is to lend its support to the operationof human rights in the information society by improving ties between the technology

community, the policymaking community, and civil society at the international level.


5/90

Ethical Implications of Emerging Technologies:

A SurveyMary Rundle and Chris Conley

Foreword 4

Introduction 6

The Technologies as a Short Story 8

Infoethics Goals for Neutral Technologies 11

The Ethical Challenges of Emerging Technologies - Case Studies 28

The Semantic Web and Other Metadata 28

Digital Identity Management 31 Biometrics 38

Radio Frequency Identification 41

Sensors 50

The Geospatial Web and Location-Based Services 55

Mesh Networking 58

Grid Computing 62

New Computing Technologies 70

Table: Summary of Infoethics Concerns 73

The Short Story Revisited 75

Recommendations 78

ANNEX: A Democratic Information Society

(Summary of an interview with David P. Reed) 82

Table of Contents


6/90

4

Embracing coherent ethical guide-lines is essential for building inclusiveknowledge societies and raisingawareness about the ethical aspectsand principles is central for upholdingthe fundamental values of freedom,equality, solidarity, tolerance andshared responsibility. Thus, UNESCOencourages the definition and adop-tion of best practices and voluntary,

professional guidelines addressingethical issues for media professionals,information producers, and serviceproviders and users with due respectto freedom of expression.

The quickening speed of techno-logical evolution leaves little time todecision-makers, legislators and other

major stakeholders to anticipate andabsorb changes before being chal-lenged to adapt to the next wave oftransformation. Lacking the time forlengthy reflection, the internationalcommunity is often faced with imme-diate policy choices that carry seriousmoral and ethical consequences:

Increase public infrastructure orpermit preferential use by investors?Allow the market to oblige peopleto participate in digital systems orsubsidize more traditional lifestyles?Let technology develop as it will orattempt to programme machines tosafeguard human rights?

The Infoethics Survey of Emerging

Technologies prepared by the NGOGeneva Net Dialogue at the requestof UNESCO aims at providing anoutlook to the ethical implications offuture communication and informa-tion technologies. The report furtheraims at alerting UNESCOs MemberStates and partners to the increasingpower and presence of emerging

technologies and draws attention totheir potential to affect the exerciseof basic human rights. Perhaps as itsmost salient deduction, the studysignals that these days all decisionmakers, developers, the corporatescholar and users are entrusted with aprofound responsibility with respect

Foreword


7/90

5

to technological developments andtheir impact on the future orientationof knowledge societies.

It is our hope that this study willimpress upon the policy makers, com-munity, producers and users the needto carefully observe evolutions in ICTs and, by so doing, to comprehendthe ethical and moral consequences

of technological choices on humanrights in the Knowledge Societies.

Abdul Waheed Khan

Assistant Director-General forCommunication and Information,UNESCO


8/90

6

Introduction

The Internet boom has providedmany benefits for society, allowing thecreation of new tools and new waysfor people to interact. As with manytechnological advances, however, theInternet has not been without nega-tive aspects. For example, it has creatednew concerns about privacy, and it hasbeen hampered by spam and viruses.Moreover, even as it serves as a medium

for communication across the globe, itthreatens to cut off people who lackaccess to it.

New solutions in information and com-munication technologies (ICTs) areconstantly emerging, and, for good orfor ill, the changes they bring may openup our societies and our world to a

greater extent than did the first phase ofthe Internet revolution. It is imperative toconsider the implications of these newtechnologies, and to encourage positivechoices regarding their uses.

UNESCO is well situated to call theattention of the international commu-

nity to these advances, particularly as totheir ethical and societal consequences,which this report refers to as infoethicalchallenges or infoethics. To that end,this survey analyzes certain UNESCOgoals in light of emerging technologiesthat will usher in the future InformationSociety1 in particular:

1. The Semantic Web and Other

Metadata Metadata, or dataabout data, enables greater auto-mated analysis of information; theSemantic Web promises to usemetadata to create an environ-ment in which computers canserve as intelligent agents ratherthan mere tools.

2. Digital Identity Managementand Biometrics Digital identitymanagement allows the amass-ing and automatic processing ofpersonal data; biometrics providesmeans by which human beingscan be uniquely identified.

1 UNESCO is increasingly using the term Knowledge Societies, which reflects a

development-oriented, people-centred and pluralistic vision of the future societies.However, for the purposes of this study, the term Information Society is used throughout

the document.


9/90

7

Introduction

3. Radio Frequency Identifica-

tion (RFID) and Sensors Thesetechnologies monitor the physi-cal world, using communicationstechnology to distribute informa-tion about a specific location.

4. The Geospatial Web and Loca-

tion-Based Services Both ofthese technologies serve to as-sociate digital data with physicallocations.

5. Mesh Networking Meshnetworking facilitates the forma-tion of networks across areaswithout existing communicationsinfrastructures. As such, it can help

connect underserved areas.

6. Grid Computing This tech-nology may allow the worldscomputing power and data stor-age resources to be pooled forpeople to access as needed.

7. New Computing Technologies

Combined with the technolo-gies listed above, a powerful mixof optics, quantum computing,and other new technologies haspotential to bring about a globalbrain.

Because choices in their design and

use carry moral consequences, thesetechnologies pose significant infoeth-ics challenges.

This survey considers these choices

in the light of key UNESCO infoethicsgoals - in particular:

(a) Fostering the application ofhuman rights and fundamentalfreedoms in cyberspace;

(b) Extending the public domain ofinformation;

(c) Enabling diversity of content ininformation networks; and

(d) Promoting access to informationand means of communication.

Taking these objectives as a given, the

survey employs them as measures inassessing likely consequences of differ-ent technological choices.

In presenting results of this examina-tion, the report first tells an introductorystory of how the technologies coveredrelate to one another. Next, infoethicsgoals are presented. Then, for each tech-

nological trend surveyed, the reportcontains a short chapter drafted in layterms to provide an overview of therelevant technology and to highlightramifications and concerns. The reportthen summarizes this infoethics analysisand revisits the story of the emergingtechnologies. Finally, the report offers

recommendations on ways to advanceinfoethics goals in anticipation of theseoncoming technologies.


10/90

8

The Technologies as a

Short Story

In the short history of the InformationSociety, technology has moved frommaking sense of cyberspace tomaking sense of the physical world,with new modes of connectivity nowholding promise for a seamlesslyintegrated Internet to reach allregions of the globe.

For the Internets first phase, humans

initiated the exchange of text, images,and other information. Given the enor-mous amount of content and codethat has been generated, computersnow need interoperable metadata,or data about data, to navigate. Thesemantic web promises to offer suchmetadata. This new metadata languageoffers predictability in a cyberspace of

ever growing exchanges, with the vo-cabulary of metadata lending greaterprecision to human use of the Internetor even allowing computers to accessand analyze content directly.

Since people have been the actorsin the first phase of the Internet, and

should remain of central concern inall future developments, it makessense that computers would need

a detailed set of terms to facilitateexchanges on behalf of individuals.Put another way, as programmesnavigate the web (with web agentsquerying many web sites to answerany given human question), a personneeds to be able to delegate hisidentity to a programme so it canthink and act on his behalf forexample negotiate for his preferred

rental car model, designate whichfriends may have access to hiscalendar, or pay taxes on onlinetransactions. Hence, records arebeing developed to refer to elementsof a persons various digital personae(e.g., name, date of birth, citizenship,etc. for one persona; pseudonym,favorite songs, etc. for another). In this

way, metadata serves as a languagefor describing digital identities.

To date, digital identities have typi-cally been dissociated from physicalidentities. However, the emergingtechnology ofbiometrics promises tocorrelate the two, linking an individuals

various digital identities to his embod-ied person. An embodied person canbe represented in digital form through


11/90

9

The Technologies as a Short Story

the translation of unique attributes for

example his fingerprints, iris pattern, orwalking gait. These attributes are takenas measurements, with the biometricstranslated into numerical expressionsthat computers can refer to in theirmachine-readable language. A physi-cal person can thereby be uniquelyidentified and then abstracted as data.

Similarly, both radio frequencyidentification (RFID) and sensorsare making other aspects of thephysical world manageable andsearchable in the digital world. RFIDtags enable the tracking of physicalobjects or people using cheap,digital technology: A person bearing

an RFID chip can easily be identifiedfor different purposes for example,ensuring that new-born babies donot get mixed up in a hospital, orgranting the proper people accessto restricted areas of a building.Likewise, a specific product (say, abottle of shampoo) can be trackedfrom its production line to the store

where it is sold and even associatedwith an individual consumer. Thisinformation can be used for a range ofpurposes, from promoting effi ciencyin the supply chain to allowing therecall of faulty products.

One might take issue with metadatas

linguistically equating an individualwith a bottle of shampoo. To such anobjection, a computer scientist mightrespond that these problems can beresolved as the metadata languagesare refined to add dimension andascribe value.2

There are certain signs that thisrefinement is underway, drivenin part by the importation of thephysical world into cyberspace.Sensors are enabling the furtherdigitization of the physical world measuring observable qualitiessuch as oxygen level or acceleration,and then translating them into digital

form. With more material added tocyberspace, flesh is put on the bonesof previously catalogued ideas.Different contexts provide dimensionand establish added meanings forthe semantic web to recognize.

Meanwhile, just as cyberspace isimporting the real world, it is itself

being exported to that physicalspace through the geospatial weband location-based services (LBS)technology. These Internet-renderedservices superimpose themselves onthe real world and offer web-basedviews of real-world locations overlaid

2 Such issues point to the problem of technological neutrality that is, the question of

whether technology is neutral or whether it is laden with values. Values may be ascribed

during the development of a language, and someone or something needs to choosehow to order things for example deciding if a human falls into the same category as an

object, or how categories should overlap.


12/90

10

The Technologies as a Short Story

with relevant information (e.g., home

prices, crime rates, or hiking trails).The line separating the real worldand cyberspace has already begunto blur; in time, these two worlds maybe integrated.

However, the Information Society isnot there yet. Huge swaths of the globeremain without Internet connectivity,particularly in the developing world.While this territory historically hasseemed vast, it may not be long beforeglobal connectivity is realized. Theadvent ofmesh networking, whichallows network-enabled devices (e.g.,mobile phones) to establish a peer-to-peer network spontaneously, offers

one means of extending the rangeof connectivity without requiring anexpensive infrastructure. The evolvingability of mobile phones and otherelectronic devices to connect tothe Internet and to each other willexpand the reach of mesh networks,perhaps eventually extending a globalnetwork to every location on earth in

ubiquitous networking.

Of course, a network comprisedlargely of devices without significantcomputing power has limitationsin terms of computing resources,data storage and accessibility. Thisproblem could be addressed by grid

computing, whereby storage andcomputing power are pooled in anetwork, with people tapping intoresources as needed and providing

them as available, in accordance with

some cost allocation scheme.

Should a combination of ubiquitousnetworking and grid computing comeabout, the Information Society wouldfind itself in a world of ubiquitouscomputing at which point thephysical and digital worlds will becomeless distinguishable. Every object even doors, clocks, refrigerators, orwatches will be able communicate itsstatus and respond to its environment.Meanwhile, computing power itselfmay continue growing exponentially,supported by new technologies suchas optical or quantum computing,which will be important for processing

this immense quantity of data. Incombination with a massive computinggrid, this force could constitute a largevirtual brain that continuously pushescomputational limits in an expandinguniverse.

This story tells of a bright future inwhich emerging technologies are

applied to the benefit of all human-ity. History suggests, however, thattechnology can also be used to limitrather than to promote human rightsand dignity. Thus, it is important toconsider how these technologiesmay promote or thwart the realiza-tion of infoethics goals.


13/90

11

Information and communicationtechnologies (ICTs) are an increasinglypowerful force in the modern world.

Their influence can be seen in allspheres of public life, from businessinteractions and education to politicsand international affairs. Thesetechnologies, particularly the Internet,have also become a dominantmechanism for conducting private

affairs and participating in society.Advancing technology frequentlyserves to allow prior activities to beperformed more easily and can openup entirely new possibilities.

Moreover, the rate at which tech-nology advances is itself increasing.Moores Law,3 which states that the

computing power of a single mi-croprocessor grows exponentially,applies to other technologies aswell. The Internet and related tech-nologies allow for new ideas andinventions to be distributed far morerapidly than before, and encouragethe ever-increasing pace of tech-

nological growth.

But technology in itself is neutral; itdoes not directly contribute to theadvancement of human rights. Manytechnologies have multiple possibleapplications, some of which mayserve to further this goal, others ofwhich may hinder it.

It is therefore imperative that emer-ging technologies be examined in

light of their impact on the exerciseof human rights. As describedabove, the infoethics goals providea framework for this examination.

The first infoethics goal, derived fromthe Universal Declaration of HumanRights,4 establishes the fundamentalpriority of putting technology in theservice of human rights. Stemming

from that goal are three others thataim to promote the public domain,diversity of content, and accessto information and the means ofcommunication with these threebased on the premise that all peopleshould be able to share in thebenefits of ICTs.

3 Seehttp://en.wikipedia.org/wiki/Moore%27s_law (viewed November 8, 2006).4 UN General Assembly resolution 217 A (III) of 10 December 1948.

Infoethics Goals for

Neutral Technologies
http://en.wikipedia.org/wiki/Moore%27s_lawhttp://en.wikipedia.org/wiki/Moore%27s_law


14/90

Infoethics Goals for Neutral Technologies

12

Human Rights and Fundamental

Freedoms

Technology can serve to promote orrestrict human rights. The InformationSociety should foster the use ofemerging technologies in such a wayas to maximize the benefits that theyprovide while minimizing the harms.In many cases, this promotion may beless a matter of technological controlthan of oversight: establishing theproper legal or regulatory system toensure that technology capable ofabuse is not in fact abused, and thatthe benefits of technology are sharedamong all.

In discussing human rights, we takeas a starting point the UniversalDeclaration of Human Rights. Manyof the rights stated in this documentare particularly important whenconsidering the ethical ramificationsof new technologies and theirpotential uses. For the purposes ofthis study, key provisions include:

ARTICLE 2:

Everyone is entitled to all the rights

and freedoms set forth in this

Declaration, without distinction of

any kind, such as race, colour, sex,

language, religion, political or other

opinion, national or social origin,property, birth or other status.

Furthermore, no distinction shall

be made on the basis of the politi-

cal, jurisdictional or international

status of the country or territory towhich a person belongs, whether

it be independent, trust, non-self-

governing or under any other

limitation of sovereignty.

ICTs enable the collection and pars-ing of information and, as such, allowmyriad categorizations. Personal datamay be broken down to such elementsas ethnicity, gender, religion, national-ity, and socio-economic status, amongothers. From an infoethics perspec-tive, it is important to ensure that thecategorization of data elements ac-cording to such lines does not result ininterference with a persons rights and

freedoms.

ARTICLE 3:

Everyone has the right to life,

liberty and security of person.

The right to life, liberty and securityof person is one of the most

fundamental rights included in theUniversal Declaration of Human Rights,and yet among the most diffi cultto define. This right encompasses aright of access to all of the necessitiesof life, including food and shelter.

ICTs contribute to health-improving

and even life-saving benefits, rangingfrom clean-air technologies andcoordinated medical research to earlyemergency alerts and quick access


15/90


13

to medical information. Does a right

to life encompass universal accessto these benefits of technology,regardless of ability to pay?

So, too, the economic effi cienciesbrought on by ICTs may raise thestandard of living in many areas ofthe world. Hence, the economicgains normally translate into betterenjoyment of the right to life, libertyand security of person.

While this Article may be construedas entailing a right of access to theinformation, ideas, cultural elements,and communication media that allowpeople to take part in society, it may

also be read to allow an individualto opt out of participating in ICTsystems. For example, under thisright a person might be permittedto refuse to have an ICT deviceimplanted into his body. If such animplant were to become de factorequirement for participation in theInformation Society, should the law

step in to allow that person to haveaccess to all of the necessities of life,including food and shelter, despitehis refusal to receive the implant?

ARTICLE 7:

All are equal before the law and

are entitled without any discrimi-nation to equal protection of the

law. All are entitled to equal pro-

tection against any discrimination

in violation of this Declaration and

against any incitement to suchdiscrimination.

The concept of equal protection canbe aided or hindered by the deploy-ment of ICTs. Technologies that serveto provide all people with equalopportunity and access, withoutdiscrimination, serve this provision.However, technology can also beimplemented to identify members ofspecific groups, and thus to enablediscrimination against those groups.

ARTICLE 11:

(1) Everyone charged with a

penal offence has the right tobe presumed innocent until

proved guilty according to

law in a public trial at which

he has had all the guarantees

necessary for his defence.

(2) No one shall be held guilty of

any penal offence on account

of any act or omission whichdid not constitute a penal

offence, under national or

international law, at the time

when it was committed. Nor

shall a heavier penalty be

imposed than the one that

was applicable at the time the

penal offence was committed.

Data that is gathered or analyzed bytechnology is increasingly used in


16/90


14

judicial proceedings. When evidence

stemming from the application oftechnology contradicts a persons tes-timony, a dilemma arises as to whichaccount deserves more credence.Despite the tendency of society to trustevidence produced by machines onthe basis of general statistical accuracy,computer code may carry mistakes orbe corrupted. In this sense, then, thereare infoethics nuances in the use oftechnology to establish proof.

ARTICLE 12:

No one shall be subjected to arbi-

trary interference with his privacy,

family, home or correspondence,

nor to attacks upon his honor andreputation. Everyone has the right

to the protection of the law against

such interference or attacks.

ICTs can serve to protect or limitthe right of privacy. For example,encryption technologies can makecommunication between private

parties confidential, or they can bemodified to allow outside interests(such as governments) to interceptthis communication. Likewise, sur-veillance technologies can serve toguard privacy, but they may also beused as tools to infringe on the pri-vacy of others.

ICTs can afford anonymity, allowing

people to feel comfortable sharingideas they might not air if their nameswere associated with these ideas. Inthis sense, privacy and anonymity incommunication bear a close relation-ship to the right to seek, receive andimpart information (Article 19) as wellas the right to associate or assemble(Article 20).

Efforts to protect privacy, however,may impose other costs on society,and efforts to protect other rights mayhave privacy implications. For exam-ple, any protection that technologyprovides for anonymous, secure com-munication may limit the effectiveness

of protection against attacks on apersons honor and reputation.

Privacy concerns are also raised bythe increasing collection of personaldata by private and governmentalentities. A question receiving grow-ing attention is the duty that privatecompanies have to safeguard con-

sumers personal data, especially in theinternational context of the Internet.

There are also questions regardinggovernmental treatment of personaldata, particularly concerning whatrules apply when different govern-ments share such information.5

5 The United States and the European Union are currently embroiled in a dispute concerning

the disclosure and use of information about airline passengers. See Air Security Talks Fail,

Daily Mail, October 2, 2006, p. 32.


17/90


15

Of course, there are questions as to

what constitutes arbitrary and in-terference. If a policy applies acrossthe board and is supported by whatseem to be reasonable arguments, isit arbitrary? If surveillance poses noinconvenience, is it interference?6

ARTICLE 18:

Everyone has the right to freedom

of thought, conscience and reli-

gion; this right includes freedom

to change his religion or belief,

and freedom, either alone or in

community with others and in

public or private, to manifest his

religion or belief in teaching, prac-

tice, worship and observance.

ICTs and thought, conscience andreligion can interact in various ways.First, and most simply, ICTs can beused to further religious interestsand to allow persons to communi-cate on issues of faith. If, however,the use of technology is contrary

to beliefs, then the observance ofthese beliefs may be threatenedby making the use of technologymandatory or practically necessaryto function in the modern world.

ARTICLE 19:

Everyone has the right to freedom

of opinion and expression; this

right includes freedom to hold

opinions without interference and

to seek, receive and impart infor-

mation and ideas through any

media and regardless of frontiers.

As with the right of privacy, theright to freedom of opinion andexpression in the Information Societyis tightly intertwined with ICTs.

Technologies can open channels bywhich information may be sharedand opinions expressed; it can alsobe used to restrict the information

available7

and to identify and interferewith people expressing alternativeopinions.8 In this sense, there is a linkbetween privacy (Article 12) and thefreedom to seek, receive and impartinformation.

Moreover, the right to freedomof opinion and expression loses

value without the ability actually tocommunicate ones views to others.ICTs can be used to create a publicforum where this communicationcan take place, or it can restrictexpression by placing limits on a

6 See Lawrence Lessig, Code and Other Laws of Cyberspace, New York: Basic Books, 1999, Ch.

11: Privacy.7 For example, many countries currently use filtering software to limit the information that

citizens can access on the Internet. Seehttp://www.opennetinitiative.org/.8 See, e.g., http://www.rsf.org/article.php3?id_article=16402 (describing the prosecution of

dissidents based on information provided by Yahoo!).
http://www.opennetinitiative.org/http://www.rsf.org/article.php3?id_article=16402http://www.rsf.org/article.php3?id_article=16402http://www.opennetinitiative.org/


18/90


16

Anonymous Expression

By Wendy Seltzer9

Anonymous expression has a long and distinguished tradition. Voltaire and George

Eliot wrote under pseudonyms. Support for the ratification of the U.S. Constitutionwas procured through anonymous articles in the Federalist Papers. Modern-day

bloggers and mailing-list contributors may not use the same flowery language and

elegant pseudonyms, but their freedom of expression is no less important. The tech-nology they use can facilitate either identification or anonymity and that will affect

the range and content of online expression.

Anonymity can make possible or enhance many expressive activities. The freedomto impart information thus includes the right to speak anonymously; freedom of as-

sembly includes the right to associate without giving a name or without revealing

group membership to outsiders; and the freedom to seek and receive informationincludes the right to listen, watch, and read privately.

Protections for anonymity are vital to democratic discourse. Allowing dissenters

to hide their identities frees them to express minority views critical to an informeddemocratic discourse. Otherwise, fear that their identity may be uncovered and that

they may be persecuted on account of their speech, may prevent those in political,

ethnic, religious, or other minority groups from speaking at all. That silence in turndeprives the whole public of access to those ideas.

persons ability to communicate with

others. The right to seek, receive,and impart information, therefore, isclosely tied to freedom of assembly

and association (Article 20) since

accessing and disseminating ideasoften entail connecting with othersin the Information Society.

9 Wendy Seltzer is Visiting Assistant Professor of Law, Brooklyn Law School, and a fellow at

the Berkman Center for Internet and Society at Harvard Law School.


19/90


17

ARTICLE 20:

(1) Everyone has the right to

freedom of peaceful assembly

and association.

(2) No one may be compelled to

belong to an association.

The right of association is affectedby ICTs in various ways. Technologymay serve to enable this right byproviding the means of facilitatingcontacts, coordinating exchanges,and interacting with other persons inan association. However, technologycan also hinder this right, if used toidentify and target members of anassociation, or to disrupt or otherwise

prevent peaceful assembly.

Technology also presents a threat tothe right to refrain from joining anassociation. This right to be alonedepends in no small part on theindividuals choice not to interactwith others. Technology can infringeupon that right, requiring that a

person associate with others in orderto obtain the full benefits available tomembers of society. It can also enablethe identification and stigmatizationof those who choose not to join agiven association.

As noted above, the rights to enjoy

privacy (Article 12) and to seek, receive,and impart information (Article 19)are linked to this package of rights

with technology reinforcing this

relationship.

ARTICLE 21:


take part in the government

of his country, directly or

through freely chosen repre-

sentatives.

(2) Everyone has the right of

equal access to public service

in his country.

(3) The will of the people shall

be the basis of the authority

of government; this will shall

be expressed in periodic andgenuine elections which shall

be by universal and equal

suffrage and shall be held by

secret vote or by equivalent

free voting procedures.

Democratic elections, like many otherfacets of modern life, are becoming

increasingly reliant on technology.Political candidates rely on media andcommunication networks to expresstheir views and coordinate sup-porters. Candidates are not alone indoing so; politically active groups arealso increasingly using the Internet toobtain grass-roots support for their

proposals and to build a constituencyand thus a voice.


20/90


18

Political use of ICTs can be harmful,

however, where technology is abusedto further a political candidacy oragenda.10 Furthermore, the increasinguse of technology in politics canpresent a barrier to entry, where certainclasses of persons are effectivelybarred from political activity if ICTsare not broadly available for theiruse. Finally, as electronic voting isadopted as a means for carrying outelections, security against corruption isincreasingly dependent on specialists.

ARTICLE 26:


education. Education shall be

free, at least in the elemen-tary and fundamental stages.

Elementary education shall

be compulsory. Technical and

professional education shall

be made generally available

and higher education shall

be equally accessible to all on

the basis of merit.

(2) Education shall be directed to

the full development of the

human personality and to the

strengthening of respect for

human rights and fundamental

freedoms. It shall promote

understanding, tolerance and

friendship among all nations,

racial or religious groups, andshall further the activities of

the United Nations for the

maintenance of peace.

(3) Parents have a prior right to

choose the kind of education

that shall be given to their

children.

Education is becoming reliant on tech-nology in two ways. First, educationabout technology itself is growingin perceived importance and valueas technology emerges as a domi-nant facet of business and thus aviable career route for many learners.

Second, technology is used to enableeducation on a vast range of subjects,allowing learners to improve theiraccess to outside sources of informa-tion, use multimedia educationalmaterials, and interact with teachersand fellow learners in new ways.

For these reasons, the single greatest

threat that ICTs pose to the right toeducation is the possibility that theywill serve to increase stratificationbased on income level and accessto technology. As ICTs become a keycomponent in an educational system,learners who are unable to obtainaccess to the technology have fewer

10 For example, staff members of several members of the U.S. Congress have altered onlineencyclopedia entries about their employer, sometimes removing facts that cast the Senator

or Representative in a negative light. See http://news.bbc.co.uk/2/hi/technology/4695376.

stm.
http://news.bbc.co.uk/2/hi/technology/4695376http://news.bbc.co.uk/2/hi/technology/4695376


21/90


19

resources available to them; moreo-

ver, the quality of non-technologicaleducational resources may decline asa result of a greater focus on onlineeducation. Encouraging programmesthat seek to prevent this result shouldbe a primary concern of governmentsand infoethicists.11

Finally, technology can impact theability of parents to choose the edu-cation presented to their children.ICTs can provide a broad rangeof options, allowing schools andparents to create individually craftedcurricula that take advantage of thewide range of options. Conversely,however, parents may find it diffi cult

to limit the information that theirchildren access as part of their educa-tion precisely because of the ease ofaccess provided by technology.

ARTICLE 27:

(1) Everyone has the right freely

to participate in the cul-

tural life of the community, toenjoy the arts and to share in

scientific advancement and

its benefits.

(2) Everyone has the right to the

protection of the moral and

material interests resulting

from any scientific, literary or

artistic production of whichhe is the author.

Perhaps the greatest promise of ICTsis the concept of a true informationcommons, a medium that allowsfor the increasingly rapid discoveryand distribution of new ideas. At thesame time, technology can serveas a barrier to the spread of newideas. For example, ideas that areonly available on a specific mediummay infringe on the rights of thosewho have no access to the requiredtechnology. Furthermore, even thosewith access to the Internet and otherforms of ICTs can be blocked from

full participation in cultural life bytechnological means, for exampletechnology implementing data userestrictions.

Technology also poses a threatto existing intellectual propertyregimes and thus to the protection ofrights-holders interests. File-sharing

networks and other activities madepossible by new technology havemade the infringement of copyrightmuch easier to effectuate and harderto prevent or prosecute.

In order to encourage a vibrantcultural life for all, infoethics must

11 Two such programmes are the One Laptop per Child programme at the MassachusettsInstitute of Technology (see text box, infra, and http://laptop.media.mit.edu/) and the

Global Education and Learning Community (see http://www.sun.com/products-n-

solutions/edu/gelc/).
http://laptop.media.mit.edu/http://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://www.sun.com/products-n-solutions/edu/gelchttp://laptop.media.mit.edu/


22/90


20

consider the role that technological

advancement may play in strength-ening or weakening the enforcementof intellectual property rights.

Access to Information and

Communication

In order for human rights to befully operational in the InformationSociety, people need access to infor-mation, which in turn requires accessto the means by which it is delivered.

Therefore, infoethics goals must alsofocus on three major categories ofaccess to information and communi-cation that are essential to the exerciseof human rights: (i) the public domain

of knowledge and creative works;(ii) diversity of content on informationand communication networks; and(iii) unfettered access to such informa-tion (including the means of deliveryand the ability to use content).

Public Domain

One primary goal of infoethics isto extend the public domain ofinformation; that is, to define an illus-trative set of knowledge, information,cultural and creative works thatshould be made available to everyperson. This category contains, but isnot necessarily limited to:

Government documents, allo-wing an informed democracyto observe and evaluate the

actions of their elected leaders,

and thus allowing all personsto participate in the process ofgovernment;

Information about personal

data retained by entities, allo-wing individuals to understandthe extent to which actions mayor may not be private;

Scientific and historical data,allowing all persons full access tothe knowledge that they need tointerpret events and to furtherthe progress of knowledge;

Information relating to health

hazards, allowing persons tounderstand the risks to whichthey may be exposed and to actaccordingly;

Information on the state

of technology, allowing thepublic to consider how theInformation Society might guard

against information warfare andother threats to human rights.

Creative works that are part

of a shared cultural base, al-lowing persons to participateactively in their community andcultural history.

UNESCOs Recommendation concerningthe Promotion and Use of Multilingualism

and Universal Access to Cyberspace,


23/90


21

adopted at its 32nd session in October

2003, provides the following definitionof public domain information: Publicdomain information is publicly acces-sible information, the use of whichdoes not infringe any legal right, orany obligation of confidentiality. It thusrefers on the one hand to the realm ofall works or objects of related rights,which can be exploited by everybodywithout any authorization, for instancebecause protection is not grantedunder national or international law, orbecause of the expiration of the term ofprotection. It refers on the other handto public data and offi cial informationproduced and voluntarily made avail-able by governments or international

organizations.

The extent of the public domain isfrequently contested. Public accessto government documents and de-liberations is naturally restricted bythe need for confidentiality in certainaffairs; similarly, access to personalinformation is constrained by privacy

concerns. Some nations choose to

remove certain content from the

public domain, deeming exposure tothis material harmful to the popula-tion at large.12 Scientific research andknowledge may be limited by govern-ment regulation, often due to ethicalconcerns.13 Intellectual property lawsfrequently reduce the public domainby granting exclusive license overcreative works to the holder of the in-tellectual property rights, with varyingdegrees of fair use permitted.14

However, infoethics perhaps neednot determine whether any given linebetween public domain and restrictedinformation or intellectual property isright. Instead, infoethics should focus

on ensuring that information that isclearly part of the public domain isavailable to all persons. Information onhealth risks of new technologies shouldbe readily available and distributed toall potential users. Creative works thatare part of the public domain should beclearly indicated as such. Governmentsshould allow access to documents that

are not properly secret,15 including

12 In Germany, for example, content promoting neo-Nazi organizations or denying the

Holocaust is prohibited. See Deutsche Welle, Trial Highlights Limits of Free Speech in

Germany, available at http://www.dw-world.de/dw/article/0,2144,1896750,00.html

(describing the criminal trial of Ernst Zndel on charges of inciting racial hatred basedon his denial of the Holocaust).

13 For example, the UN has recently passed a resolution urging Member States to prohibit

human cloning in any form, but it was unable to reach an agreement to pass a binding formof the same resolution. See United Nations Declaration on Human Cloning, U.N. Doc. No.

A/59/516/Add.1 (2005).14 See Ruth Okediji, Towards an International Fair Use Doctrine, 39 Columbia J. Transnat L.

75 (2000).15 South Africa provides individuals a right of action to obtain information held by private entities.

See Promotion of Access to Information Act, Act No. 2 of 2000 (The Republic of South Africa).
http://www.dw-world.de/dw/article/0,2144,1896750,00.htmlhttp://www.dw-world.de/dw/article/0,2144,1896750,00.html


24/90


22

making those documents available

over commonly used communicationand information media.

Diversity of Content on

Information Networks

ICTs also can have a great impact onthe diversity of content on informa-tion networks. In an ideal society,the content available on informationnetworks should reflect the diver-sity of legitimate preferences of thepopulation. In addition, the informa-tion networks should be open tocontent from all sources, allowingany interested person to be a creatorof content rather than a mere con-

sumer.

Broadcast media, such as televisionand radio, allow content to be rap-idly delivered to consumers in distantplaces. However, broadcast mediatends to cater to the populationsegments with the most economicpower; the substantial startup costs

of operating a television channel orradio station deter the distribution ofcontent targeted at niche audiences,16

and existing channels may have no

incentive to accept content fromother sources or to provide diversecontent.17

In Brazil, observers have expressedconcern about the diversity ofcontent available on cellular phonenetworks.18 Cellular phone networksare currently the dominant means ofdistributing interactive content; while45% of the population has access tocellular phones, only 12% has accessto the Internet. Not surprisingly,then, cellular phone networks areincreasingly becoming carriers of var-ious forms of content beyond voicetelephony; music, video, interactive

games, and other material can beaccessed on cellular phone networks.

Thus, as with traditional media enter-prises, cellular phone companiesmay have exclusive control over thecontent available to a large fractionof the population, and therefore havethe potential to promote or limit thedelivery of available content.

ICTs offer the potential to overcomethese obstacles to content diversity.

16 Some nations have attempted to address this problem through regulation. In Germany,for example, each state must either provide a public network supplying diverse content or

must regulate private networks to ensure that they provide a diversity of content matched

to the interests of subsets of the population. See Uli Widmaier, German BroadcastRegulation: A Model for a New First Amendment?, 21 B.C. Intl & Comp. L. Rev. 75, 93-99

(1998).17 The United States attempted to regulate the relationship between television networks and

the creators of television programming, but has abandoned that effort. See Christopher S.

Yoo, Beyond Network Neutrality, 19 Harvard J. L. & Tech. 1, Fall 2005, p. 49 n.188.18 Email conversation with Ronaldo Lemos, Creative Commons Brazil, February 2006.


25/90


23

In principle, ICTs substantially reduce

the cost of producing and distribut-ing content, allowing diversity ofboth creative works and creators.In allowing the generation of moreinteractive content, they transformconsumers into active participants.

Unfettered Access to

Information

Establishing a base of public domaininformation and diversity of contenton information networks are laud-able goals. However, the fulfillmentof the spirit of these goals dependson the actual ability of persons toaccess and use the content and infor-

mation that the public domain anddiversity make available. Thus, oneof the dominant goals of infoethicsis to achieve universal access to alllegal content. There are two, largelydistinct, components to this goal:ensuring that all persons are able toobtain the content, and ensuring thatall persons are able to use the content

that they obtain.

The ability to obtain content requiresaccess to the information networksor other channels of distribution,which can be accomplished eitherby increasing access to a particularchannel or by providing additional

channels of distribution. For instance,

a country whose citizens have limited

access to computers and the Internetcannot achieve the goal of universalaccess to information simply byensuring that content is availableonline. Instead, it must determine thebest course of actually making thatcontent available to its population.

This might be achieved by deployingnew technologies like web-enabledmobile phones to increase connec-tivity; it might also be accomplishedby ensuring that the content is alsoavailable in other forms to those whocannot access it online.

Diversity of content is heavily relianton maintaining network neutrality

so that no single entity can serve tolimit access to (legal) content.19 Undernetwork neutrality, each node on thenetwork is content-blind, passingalong all traffi c without concern asto its type or content. Thus, if fully im-plemented, network neutrality wouldrequire that all nodes on a network(including Internet service providers,

or ISPs) pass along traffi c withoutconcern for (or even identification of)its origin, destination, and content.

Complete neutrality, however, carriesits own hazards: a network that isentirely content-neutral encouragesthe diversity of legal content, but it

also permits various forms of illegal

19 See, e.g., Mark N. Cooper ed., Open Architecture as Communications Policy, Center for

Internet and Society at Stanford Law School, 2004.


26/90


24

content, ranging from pornography

to spam and viruses. Infoethics shouldthus seek to shape the future of ICTsto retain the positive features ofnetwork neutrality while encouragingdevelopments that limit its harms.20A key question here is whether edgedevices might better address harmsthan could changes to the networkitself since edge devices may bebetter technically and are more likelyto rest under the users control.

Search engines present another accesspoint where the practical accessibilityof content may be limited. Given thesheer volume of data available on theInternet, many users rely on search en-

gines to locate desired content; thus,any content that cannot be accessedvia search engines may be effectivelyinaccessible. Like ISPs, search enginesthus have the potential to act as achokepoint and to affect the diversityof content available on an informationnetwork, promoting some contentby placing it atop a list of search

results while selectively filteringother content. Google, for example,has filtered searches conducted onits French and German language

sites, removing matches that link to

Nazi or anti-Semitic content.21 Again,however, this capability is not ethicallyquestionable per se; it may simplyrequire oversight to ensure that it isbeing utilized to remove only illegiti-mate content.22

If information is distributed via ICTs,access to it fundamentally requiresaccess to the technology itself.Increasingly, projects are beinglaunched to remedy lack of ICTsaccess, providing a solution that maybe more cost-effective and morebeneficial than projects to provideinformation in various offl ine formats.

The One Laptop per Child Foundation,

for instance, seeks to leverage ad-vancing technology to producecheap, low-powered computers withwireless network capacity that can bewidely distributed in poorer segmentsof the world.23 (See Text Box, below.)Moreover, as will be discussed in thesections that follow, the problem ofexclusion could diminish as the cost

of networked computing drops.

Still, even if the cost of ICTs declinesto a point where todays technology

20 See Jonathan Zittrain, Without a Net, in Legal Affairs (January - February 2006), available at

http://www.legalaffairs.org/issues/January-February-2006/feature_zittrain_janfeb06.msp.21 See http://cyber.law.harvard.edu/filtering/google/results1.html; cf. Isabell Rorive,

Freedom of Expression in the Information Society, Working Paper for the Preparatory

Group on Human Rights, the Rule of Law and the Information Society 8, Sept. 15, 2004

(discussing censorship by search engines in France, Germany, and China).22 In addition, users could configure personal filters to prevent exposure to undesired

content.23 Seehttp://laptop.org/.
http://www.legalaff/http://cyber.law.harvard.edu/fihttp://laptop.org/http://laptop.org/http://cyber.law.harvard.edu/fihttp://www.legalaff/


27/90


25

becomes affordable for all, there will

always be new advancements thatbegin with limited accessibility. Inthis regard, the process of ICTs diffu-sion warrants attention.

In the development of informationnetworks, it is important to ensurethat standards organizations andthe like are not overly influencedby particular agendas. Allowing anyone group to capture a regulatory orstandard-setting organization preventsthese organizations from achieving abalance between interests. Designersof network technologies could bemandated or given incentives to offerthe greatest overall benefit to society,

rather than to support the demands ofa specific group (e.g., owners of copy-righted content).

Ensuring that all persons are able

to use the content that they obtainposes other challenges. Even wherecontent is available, it is of little valueunless it can in fact be understoodand used. Thus, universal access toinformation requires that contentbe distributed multilingually, or thattechnology be deployed to translatecontent into a usable form. Similarly,technology can be used to make infor-mation accessible to the physicallydisabled. In addition, content shouldbe easily machine-readable; publicdomain content in particular shouldbe released in a format that is broadlyused and does not require specificapplications or devices to access.


28/90


26

One Laptop per Child: The $100 LaptopBy Samuel Klein

The One Laptop per Child (OLPC) initiative, conceived at the MIT Media Lab and firstannounced in January 2005, is an effort to mass-produce cheap, durable laptops and to

distribute them throughout the world to improve education for children. The stated goal

of the project is to provide children around the world with new opportunities to explore,

experiment, and express themselves. If all goes as planned, these opportunities will beprovided entirely through the laptops, which, equipped with free software, will be tools for

creating and receiving content.

The project to date has focused on developing cheap, durable hardware designs, particularlya low-cost display with a target production cost of $100 per laptop in its early stages (less

later) and building a network of partners to help produce the necessary hardware and

software. The project relies on economies of scale: The production schedule calls for at least5 million laptops. The laptop will have built-in wireless capability, will work as part of a local

mesh network when there is no access to the global Internet, and will support innovative

power sources, including winding by hand.

The laptops are meant to be widely distributed within a given area, one to every child

in a school or region. Distribution is to be carried out through schools via national

governments. The OLPC Chairman, Nicholas Negroponte, says the team has had initial

discussions with offi cials in China, India, Brazil, Argentina, Egypt, Nigeria, and Thailand;24the team recognizes, however, that unless a country can provide one laptop for every

child in its population, hard decisions about where and how to distribute the laptopsmust be made.

There are philosophical goals to the project, in addition to the technical ones. Laptops

were chosen in part because they can be taken home and can engage the whole family.A stated intent is that children will own their laptops, though these may be among the

most expensive and novel personal items in their neighborhood.

24 From the One Laptop Per Child FAQ: http://laptop.org/faq.en_US.html (viewed November

8, 2006).
http://laptop.org/faq.en_US.htmlhttp://laptop.org/faq.en_US.html


29/90

27

The development of a deployment and education plan from the design of software

and content to be included on the machine, to research and suggestions on how toimplement an effective One Laptop per Child teaching environment is one of the morerecent project goals to take form. OLPC is currently soliciting input on how to proceed

with and study this goal.25 While OLPC is an educational project, this effort is not simply

a question of pedagogy. Distributing millions of laptops to areas where computers arescarce, and providing every member of a communitys youngest generation with a

world view and with exotic tools and knowledge completely foreign to their elders,

disrupts the status quo in a significant way.

The question of how to select content, distribute the laptops, and recommend their uses,

so as to produce desired change without unwanted social and cultural upheaval, is a crucial

question and one on which many groups may wish to engage. For an initiative of its size, OLPCis unusually open to suggestions; the project has a publicly editable list of tasks that includes

a request for panels of thinkers to study some of these issues and propose improvements

and recommendations.26

25 From http://wiki.laptop.org/wiki/OLPC_software_task_list#Eductional_community_engagement

(viewed November 8, 2006): We propose including other intellectuals, artists, civic leaders in order

to provide a diversity of experience and expertise.26 Id., http://wiki.laptop.org/wiki/OLPC_software_task_list#Strategic_research (viewed November

8, 2006).
http://wiki.laptop.org/wiki/OLPC_software_task_list#Eductional_community_engagementhttp://wiki.laptop.org/wiki/OLPC_software_task_list#Strategic_researchhttp://wiki.laptop.org/wiki/OLPC_software_task_list#Strategic_researchhttp://wiki.laptop.org/wiki/OLPC_software_task_list#Eductional_community_engagement


30/90

28

With the increasing importance of ICTsin the world comes a growing needto recognize the ethical ramificationsof new technologies. Moreover, therapid rate of technological changedemands that we understand emer-ging technologies and their potentialeffects as they are being developed,and not wait until the consequencesare manifest before we prepare for

them. By understanding tomorrowstechnologies in light of infoethicsgoals, society can better anticipatetheir effects and deploy them in amanner that leverages their benefitswhile mitigating potential harms.

The following case studies highlightsome of these technologies and flag

many of the infoethics concerns thataccompany them.

The Semantic Weband Other Metadata

What the Semantic Web Is

The Internet was conceived as amechanism to allow humans to initi-

ate the exchange of text, images, andother information. With the exponentialgrowth of content available on theInternet, however, this is increasinglybecoming impractical. Search enginesseek to mitigate this problem by provi-ding a tool to navigate the web, butthey provide only a partial solution. Tomake the web fully navigable, interop-

erable metadata, or data about data, isrequired. This metadata can also serve tomake the web more machine-readable,allowing computers to evolve fromdumb tools to intelligent agents. Thesemantic web27 promises to offer suchmetadata.

The Ethical Challenges of

Emerging Technologies -

Case Studies

27 The Semantic Web is an offi cial project of the World Wide Web Consortium, which was

founded by Tim Berners-Lee, the webs inventor.


31/90

The Ethical Challenges of Emerging Technologies - Case Studies

29

How the Semantic Web

Works

The Internet existed for nearly threedecades before it took off as a popularmedium for information and commu-nication.28 Although the Internet wasbased on common computer lan-guages, or code, from the start (TCP/IP,SMTP, etc.), what triggered its uptakewas the royalty-free nature of the lan-guages of the World Wide Web (web) that is, HTML and HTTP and thefact that HTML was particularly user-friendly. These two languages allowedthe loose coupling of machinesinvolved in exchanging information meaning any web client seeking

information could talk to any webserver, which could then providethat information remotely, in a formthat people could enjoy. Internetusage increased dramatically, and thisphenomenon spurred the creation ofmore content, which in turn gave riseto more exchange.

Given the high volume of web content,the semantic web is being designedto create a universal medium forthe exchange of data29 using thesame loose coupling properties forprogrammatic data as there was forhuman-rendered data with HTML.

This new language should provide

predictability in a cyberspace ofever growing exchanges, with thevocabulary of metadata lendinggreater precision as computers accessand analyze content directly.

The semantic web combines a setof computer languages30 to providemachine-readable descriptions ofweb content. This metadata may becreated by humans or computers,and is designed to provide contextabout the content without requiringa person or machine to actually parsethe content. Once a piece of informa-tion is tagged, the semantic web canreason about it and develop contex-

tual meanings based on observationsabout connections that the pieceof information has with others. Thisenables machines to search web sitesand carry out tasks in a standardizedway.

Although the name refers to web, thisinitiative is geared toward enabling

machines to handle data among arange of Internet applications.

28 See History of the Internet (Wikipedia entry), http://en.wikipedia.org/wiki/Internet_history

(viewed November 8, 2006).29 Seehttp://www.w3.org/Consortium/activities#SemanticWebActivity.30 For example, Resource Description Framework (RDF), Web Ontology Language (OWL),

and Extensible Markup Language (XML).
http://en.wikipedia.org/wiki/Internet_historyhttp://www.w3.org/Consortium/activities#SemanticWebActivityhttp://www.w3.org/Consortium/activities#SemanticWebActivityhttp://en.wikipedia.org/wiki/Internet_history


32/90


30

Ramifications and Concerns

The wealth of content available oninformation networks, particularly theInternet, is useful only if people canactually find and access the informa-tion that they need. The semanticweb allows people to use computersas agents to search for appropriatecontent based on a wide range ofcriteria which could include thepublic domain or intellectual propertystatus of the content, alternate sourcesof the content in different formats orlanguages, or even the existence ofevidence serving to refute the viewoffered in the content.

The wealth of content is closelyrelated to the Internet explosion,which is often credited to the networkneutrality principle that holds thatall traffi c should be treated equally.Oddly, the semantic web could cutagainst this neutrality by equippingparties with tools to filter Internetcontent based on its associated

metadata: ISPs, routers, or searchengines could use the metadata todistinguish between types of contentand grant preferential treatment tocertain traffi c, raising barriers to entry

for new service or content providers.

In this regard, the semantic websmachine-readable labels could markcontent for discrimination and reducethe ability of users to generate andshare material.

Moreover, some would contend that,in giving users the ability to accessonly the content that they desire, thesemantic web could damage publicdiscourse. The theory here is that fullparticipation in society requires aforum in which a person can make hisvoice heard, but that the semantic weband other technologies allow otherusers fully to customize their experi-ences and to receive only the content

that they explicitly request. In otherwords, the semantic web enablesend-user insularity and so indirectlydestroys the forum. Here again, theworry is that the semantic web couldin fact harm the very connectivity thatit was designed to promote.31

More theoretically, machines must

be programmed to categorize andassign values to information so that,for example, personal data can bedistinguished from weather patternsand flagged as warranting privacy. In

31 See Cass R. Sunstein, The Daily We, Boston Review (Summer 2000), available athttp://www.bostonreview.net/BR26.3/sunstein.html. Others would argue, however, that even a

seemingly low rate of exposure to differing views via the Net (e.g., 15 per cent) may suggest

greater public dialogue than had previously been the case. See work by Eszter Hargittai,Cross-Ideological Conversations among Bloggers, http://crookedtimber.org/2005/05/25/

cross-ideological-conversations-among-bloggers/ visited November 8, 2006 (describing

work by Eszter Hargittai, Jason Gallo and Sean Zehnder analyzing cross-linkages among

liberal and conservative political blogs).
http://www.bostonreview.net/BR26.3/sunstein.htmlhttp://www.bostonreview.net/BR26.3/sunstein.htmlhttp://crookedtimber.org/2005/05/25http://crookedtimber.org/2005/05/25http://www.bostonreview.net/BR26.3/sunstein.htmlhttp://www.bostonreview.net/BR26.3/sunstein.html


33/90


31

this sense, there are risks inherent in

designing tools simply to exchangeinformation without simultaneouslycoding them to assign a highervalue to data relating to humans.Safeguarding human rights mayrequire programming computers toput personal data on a higher plane.

Still, one should not overstate thedangers of the semantic web. Afterall, these harms are possible evenwithout metadata and are far fromcertain to occur even with metadata.

On balance, it seems the semanticweb will support the goal of promot-ing access to information by making

existing content far easier to identify,locate, and use.

Digital Identity

Management

The previous case study discussedhow metadata is allowing increasinglysophisticated machine-to-machineexchanges. This increase in commu-nication between machines createsthe potential for good or harmful con-sequences, such as the lowering of

transaction costs in commerce or thelaunching of malicious virus attacks.

This section first looks at how the

automated exchange of data isdriving the need for digital identitymanagement tools that afford bettercontrol over the flow of personalinformation. It sets the stage for thenext case study, which explores how,through biometrics, metadata cango beyond a persons diverse digitalpersonae to pinpoint an embodiedperson.

What Digital Identity

Management Is

Simply stated, digital identity man-agement concerns the control of

digitized information pertaining toa person. This information is some-times referred to as personal data, orpersonally identifiable information.

This latter term more precisely sug-gests that the data can be linked tothe specific individual involved.

As originally designed, the architecture

of the Internet did not provide a mech-anism to verify, or authenticate, theidentities of users. Its designers wereworking in a different time and culturefrom todays online environment, withthat early community of Internet userscomprising essentially a highly coop-erative, high-trust society of computer

scientists. The Internet they brought tobear reflected this culture.


34/90


32

With the explosion in Internet

usage drawing strangers to interactat unprecedented rates, it is notsurprising that the once trustfulatmosphere of the Internet haschanged, and that people are startingto view the space with growingreserve. Put another way, the Internethas undergone a sort of urbanization,where more and more people aregravitating toward it for its benefits,but where the traditional sense ofcommunity has broken down andpeople find they must be on guard.Computer scientists are starting tosay: In retrospect, we should havedesigned an authentication layerinto the Internet. Now, with the

Internet scaled up large, and with somuch commerce passing over it, thepotential for fraud is enormous.32

E-commerce statistics bear witness tothis shift. Statistics published last yearshowed a sharp drop in the number

of consumers who feel comfort-

able participating in e-commerce.33People have learned to questionwhether the person or entity at theother end of a transaction is indeedwho he, she, or it claims to be, andpeople wonder if they can actuallyhold that other party accountableshould something in the exchangego awry. By responding to an emailor filling out an online form, will aperson be a victim of phishing orpharming?34

At the same time, it is natural forInternet users to lament the fact thatthey currently have to rememberpasswords and fill out all sorts of forms

when reserving a rental car, purchasinga book, or engaging in some othercommon transaction online. Althoughthis is a repetitive and often frustratingtask, people have been trained toprovide information in this way, anddo so without thinking.35

32 Interview with Paul Trevithick, Project Lead of the Eclipse Foundations Higgins Trust

Framework project, August 2005.33 Riva Richmond, Internet Scams, Breaches Drive Buyers Off the Web, Survey Finds, Wall

Street Journal, June 23, 2005, p. B3, reporting on a Gartner study of 5000 online consumers.

The article states that more than 42% of online shoppers and 28% of those who bank

online are cutting back on these activities due to security and privacy issues.34 David Bank and Riva Richmond explain in Information Security: Where the Dangers Are,

Wall Street Journal, July 18, 2005: In phishing scams, fraudsters send emails that appear

to come from a trusted source, like Citibank or eBay. Click on a link in the email, and youredirected to a fake Web site, where youre asked to reveal account numbers, passwords

and other private information Then theres pharming, where hackers attack the server

computers where legitimate Web sites are housed. Type in the address of the legitimatesite, and you are redirected to a look-alike.

35 This point is often made by Kim Cameron, Microsofts Identity and Access Architect. He

reminds industry colleagues that people are not stupid, but rather they have been given

poor tools for interacting online.


35/90


33

Industry specialists say that digital iden-

tity management tools will allow onlineexchanges to be much more secureand convenient, as this technology willenable enhanced control over digitizedinformation relating to a person.36

Consumers have to date not em-braced this technology because ithas not yet been presented in a formthey are willing to accept. Microsoftis still smarting from its past experi-ence trying to provide such tools,37and other technology companieshave taken note: People do not wanta single, powerful company to be atthe center of their trust relationshipsor to occupy a monopoly position in

handling their personal data.

Technology developers38 are there-fore now focusing on user-centricapproaches to digital identity man-agement. In this new paradigm, aperson will choose among differentidentity providers to take care of hispersonal data, with his permission

then required for that identity providerto pass his information to anotherperson or entity in a transaction;

meanwhile, the new system will also

verify the identity of that other partyto the transaction.

How Digital Identity

Management Works

This user-centric identity model hastwo primary parts: one that handlesthe exchange of identity informa-tion as it passes between endpointcomputers or devices, and the otherthat helps a user manage his iden-tity information on his computer. Forthe part between devices, this newsystem may be thought of as a setof rules for exchanging information39

(called computer protocols) in theform of packaged, sealed tokens.

The user can hire whatever identityprovider he wants to guarantee hisinformation and package it intotokens, even employing different pro-viders for different purposes (e.g., oneto handle credit card details; anotherto manage core personal information

like name and date of birth; anotherto process medical records; etc.).Identity providers may reside on a

36 See, e.g., Kim Camerons Identity Weblog at http://www.identityblog.com/, Dick Hardts

Identity 2.0 weblog at http://www.identity20.com/, and others available at http://www.

identitygang.org/individuals.37 In the past several years the market has largely rejected Microsofts Passport identity

management systems. The market also resoundingly rejected Passports predecessor,

Hailstorm.38 OpenID, Sxip, the Liberty Alliance, Shibboleth, Passel, and other industry players have joined

Microsoft in the quest to provide identity management tools that the market will accept.39 A consortium involving Microsoft, IBM, and other technology firms developed the

standards for this exchange as part of a larger set of standards for web services.
http://www.identityblog.com/http://www.identity20.com/http://www/http://www/http://www.identity20.com/http://www.identityblog.com/


36/90


34

persons computer or device, or they

may be located elsewhere, accessiblethrough the Internet.

So, for example, when a person wantsto conduct an online transaction, thecomputer or device on the otherpartys end (in industry parlance, therelying party) will indicate to hisagent what package of informationis needed. The agent will then requesta token containing these claims fromone or more identity providers thatthe person has entrusted with thisinformation; the identity provider(s)will then pass the token to the relyingparty. A person can supervise thisexchange every time it occurs, or he

can do so once and then opt to let theexchange take place automaticallyhenceforth.

The second part of the system is whattakes place on a persons computeror device. Instead of rememberingpasswords or typing in an array ofinformation in an online order form, a

person conducting e-commerce willsimply choose a visual representationor icon of the particular packageneeded (e.g., an icon symbolizingbanking, medical, or tax information).When he chooses that icon, his agentinitiates a call for the release of digitaltokens by identity providers to the

relying party, as described above.The agent will be able to operate onall sorts of devices, be they desktop

computers or cellular phones or other

mobile devices. This agent will be thesole component of the digital identitysystem with which the user will needto authenticate himself directly (e.g.,through a fingerprint scan).

As the trusted intermediary, a personsidentity agent will sit at the center ofthe users communication and haveaccess to all identity informationexchanged. It will unwrap a token andtranslate the claims from one systemslanguage into a format recognizableto another. To protect privacy insofaras possible, this trusted intermediarywill ideally keep to a minimum theamount of information disclosed for

a given transaction. In many cases,this may require that the informationin a given token be transformed intoan alternate token correspondingwith a specific request. For example,the intermediary will be able to takeinformation from a token vouchingthat a person was born on a specificdate (e.g., July 20, 1969) and translate

it so that the new claim revealsnothing more than that the person isindeed over 21 years old.

Putting theory into practice, Microsoftis planning to roll out a user-friendlytoken exchange system with visualicons called Cardspace that will

resemble cards that people currentlycarry in their wallets, such as a driverslicense, a credit card, etc. While the


37/90


35

plan is to introduce this system in

the new version of Windows, calledWindows Vista,40 the digital identitysystem will also be available in up-grades for Windows XP. Microsoft hasbeen on the private campaign trailto convince big e-commerce playerslike Amazon and eBay to accept thesenew services in exchange for moredirect access to Microsoft custom-ers. Because so many people alreadyuse Windows XP, the spread of theseservices will not hinge on widespreadadoption of the newer Windows Vista.Given the fact that Windows XP runson hundreds of millions of machinestoday, these digital identity manage-ment tools have a strong probability

of taking root.

Meanwhile, IBM and Novell in February2006 announced their intention tooffer programming code to allow forsimilar digital identity managementtools to be built using open-sourcesoftware. The project, named Higgins,will enable different identity manage-

ment tools to interoperate.41 Ratherthan managing digital identities itself,Higgins overlays different systemsin order that information might beexchanged among them at a users

request.42 One attractive prospect

is portability in reputation systems,which would allow, for example, aperson to take the reputation he hasbuilt up in the eBay43 community andcarry it over into the world of SecondLife.44 The technical requirements forsuch portability are just starting to beexplored.

Ramifications and Concerns

The new digital identity manage-ment tools promise to cut todaysphishing and pharming and may alsoaddress spam problems. Since anidentity agent can help to minimize

data disclosed to a merchant or otherentity with which a person interacts,the technology may boost privacy asit minimizes the number of entitiesthat have access to an individualsprofile. Perhaps most significantly,the systems distributed architectureshould reduce vulnerability to attacksince theoretically data is not con-

centrated in one place.

If the system can protect personaldata in this way, digital identitymanagement has many possible ben-

40 Worldwide availability of Windows Vista is scheduled for early 2007.41 The Eclipse Foundation, an open-source community, manages Higgins.42 Interviews with John Clippinger, a senior fellow at the Berkman Center for Internet and

Society at Harvard Law School, Fall 2005.43 See http://pages.ebay.com/services/forum/feedback.html describing the eBay reputation

system (viewed November 8, 2006).44 Seehttp://secondlife.com/whatis/describing the 3-D virtual world of Second Life (viewed

November 8, 2006).
http://pages.ebay.com/services/forum/feedback.htmlhttp://secondlife.com/whatishttp://secondlife.com/whatishttp://secondlife.com/whatishttp://pages.ebay.com/services/forum/feedback.html


38/90


36

efits to society, including preventing

malicious conduct and making theInternet a better forum for commerce.From this vantage point, the techno-logy promotes privacy, security, andimproved living standards.

So, too, tools enabling the exchangeof personal data based on an indi-viduals preferences may facilitatesocial interactions. Paul Trevithick,Project Lead for Higgins, emphasizesthe benefits of a user-centered net-working layer that gives people morecontrol over their digital identitiesacross a wide variety of computer-mediated contexts (e.g., email, instantmessaging, e-commerce, shared

spaces, and enterprise directories),especially those involving social net-works.45 In this way, digital identitymanagement tools may serve as aboon to freedom of assembly.

Looked at from another perspective,however, the existence of digitalidentity management systems could

pose significant risks for privacy andsecurity. As noted, in the proposedarchitecture a persons identity agentwill serve as the most trusted inter-mediary in the new digital identitysystem; however, current technologyprovides no guarantee that a personsidentity agent will not collude with

the other parties to a transaction (i.e.

identity providers and relying parties).

In addition, it is quite conceivable thatthe market will not support the hostof identity providers that advocatesof systems speak of, but that insteadthere will be a natural concentrationin identity provisioning. Quite simply,users might find it inconvenient orexpensive to separate their data anddesignate elements to different iden-tity providers. Or, relying parties mightbe restrictive in recognizing identityproviders, with the result that a lim-ited set of identity providers woulddominate the market. Either way, asmall number of identity providers

would have control over a great dealof personal data. Moreover, givencurrent designs of the system, it istechnically possible for identity pro-viders and relying parties to collude.In other words, it is unclear how theplayers in the identity managementinfrastructure will be accountable tousers and the Information Society

generally.46 Of course, the market mayspur improved technologies whosearchitecture guarantees trustworthybehavior, and the law may well rein-force these incentives.47

To some, the primary infoethics con-cern is what happens if a big player

45 Interview with Paul Trevithick, September 20, 2005.46 Mary Rundle and Ben Laurie, Identity Management as a Cybersecurity Case Study,

Berkman Center Publication Series, September 2005, p. 8.47 Microsoft took a strong stance advocating privacy legislation in 2006.


39/90


37

e.g., a government or a mega-

corporation usurps a distributed,user-centric identity system and ap-plies its own globally unique identifier,violating all the user-centric principlesthat industry developers have ad-vertised. In other words, there arecurrently no innate technological pro-tections regarding possible future usesof these tools. If the tools were usedthroughout the Information Societyin an abusive manner to discriminate,intimidate, and block communication,human rights and related infoethicsgoals would be in serious jeopardy.

The magnitude of these effectsshould not be underestimated given

the revolution that digital identitymanagement may bring in machine-to-machine interactions, or webservices. Dale Olds, a key Novellengineer in this area, has indicatedthat Higgins may seek to developan additional technology that wouldautomatically transfer informationfrom an individuals given digital

identity (or persona) when he visits aweb site.48 It may be here that digitalidentity management tools will havetheir strongest impact: By enablingmachines to exchange personal dataautomatically on peoples behalf, thetools will throw off one of the biggestimpediments to web services. Hence,

the empowering of machines in this

way could ignite an explosion inmachine-to-machine interaction.

Machine empowerment points toperhaps the biggest unknown andthe one with the greatest potentialimpact concerning not what humansmay do with these tools, but ratherhow machines will treat humans withpersonal data so well organized. KimCameron, Microsofts chief Identityand Access Architect, underscoresthese apprehensions:

The broader aspects of the waynetwork intelligence respondsto who we are is of much more

concern to me when I thinkforward twenty years Beyondthe abuse of power there areother equally chilling possiblefutures involving the potentialrelationship between humankind and machine intelligence.I realize people are not likely towant to discuss this because

it is too forward thinking, butthe two actually can mutuallyreinf

Ethical Implications of Emerging Technologies - Unesco 2007

Documents

Transcript of Ethical Implications of Emerging Technologies - Unesco 2007