Virtuality and capabilities in a world of ambient intelligence

Law, Governance and Technology Series 32Luiz Costa Virtuality and Capabilities in a World of Ambient Intelligence New Challenges to Privacy and Data Protection... Since both the develo

Law, Governance and Technology Series 32

Luiz Costa

Virtuality and

Capabilities in a

World of Ambient Intelligence

New Challenges to Privacy and Data Protection

Law, Governance and Technology Series

Volume 32

Series editor

Serge Gutwirth

Brussel , Belgium

manuscripts that focus upon issues that engage into an analysis or refl exion related

to the consequences of scientifi c and technological developments upon the private sphere, the personal autonomy and the self-construction of humans with data protection and privacy as anchor points The objective is to publish both disciplinary, multidisciplinary and interdisciplinary works on questions that relate to experiences and phenomena that can or could be covered by legal concepts stemming from the law regarding the protection of privacy and/or the processing of personal data Since both the development of science and technology, and in particular information technology (ambient intelligence, robotics, artifi cial intelligence, knowledge discovery, data mining, surveillance, etc.), and the law on privacy and data protection are in constant frenetic mood of change (as is clear from the many legal confl icts and reforms at hand), we have the ambition to reassemble a series of highly contemporary and forward-looking books, wherein cutting edge issues are analytically, conceptually and prospectively presented

More information about this series at http://www.springer.com/series/8808

ISSN 2352-1902 ISSN 2352-1910 (electronic)

Law, Governance and Technology Series

ISBN 978-3-319-39197-7 ISBN 978-3-319-39198-4 (eBook)

DOI 10.1007/978-3-319-39198-4

Library of Congress Control Number: 2016949387

© Springer International Publishing Switzerland 2016

This work is subject to copyright All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfi lms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed

The use of general descriptive names, registered names, trademarks, service marks, etc in this publication does not imply, even in the absence of a specifi c statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use

The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors

or omissions that may have been made

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland

Faculté de Droit (Visiting Researcher)

University of Namur, CRIDS

Namur , Belgium

To Ribamar and Delfi na

To Raquel

Foreword

They are theses you would like to have written yourself: they fi t with your own refl ections, even if these refl ections are still immature and you are unable to express them correctly and precisely; reading through it, the thesis constitutes a clear dem-onstration or, better, an illumination of your confused ideas and anticipations

If furthermore, the thesis offers you delicious moments of intellectual adventure with its author, you feel you are the most satisfi ed man Thanks Luiz for these moments I regret not to have had more time to spend with you but I know that Professor Antoinette Rouvroy was taking over from me as I was closing my door Starting this adventure with you I had two vague convictions more than certain-ties and you accepted the task of scrutinizing them The fi rst one was my non- comfort after the decision of the EU Charter on Human Rights to separate, at least

as regards their enactment, two concepts: privacy and data protection and to neglect their deep interrelationships despite the fact that data protection might not be cor-rectly defi ned and circumvented if one doesn’t take care of its root: privacy The second one was interlinked with this fi rst one After having read Amartya Sen’s articles and books about his theory of “capabilities”, I was convinced that privacy has something to do with that theory and perhaps was a legal and ethical translation

of it I saw that assimilation as a way to reject the individualistic approach of privacy conceived more as a defence of the individuals (conceived as a liberal subject) fac-ing society’s invasion and to envisage privacy more as a concept allowing the devel-opment of our identity within a determined and democratic society and where the development of these personal capabilities within an information society is also a task for our government at the end of a deliberative process

To this extent, Luiz’s thesis has defi nitively contributed to reinforcing my victions The fi rst part of the thesis analyzes the concept of “Ambient Intelligence” and its present and foreseeable indefi nite applications By situating his refl ections

con-on artifi cial intelligence as a radical transformaticon-on of the informaticon-on and nication technologies’ social power over the individual, the author underlines the unprecedented prediction and preemption capabilities of certain actors through big data systems It underlines the normalization and the potential manipulation by cer-

commu-tain actors of human behaviour One knows the Google CEO’s assertion: ‘ It will

become very diffi cult for people to see or consume something that has not in some sense been tailored for him’ or Amazon’s: ‘ Amazon wants to ship your package before you buy it.’ Developing Rouvroy’s theory of “algorithmic governmentality”,

the author follows her thesis about the negative consequences of this ity which, due to their opacity, are making decisions incontestable under a false appearance of mathematical truth and making structural injustices less visible Perhaps Luiz’s depiction of our information society is too dark It might not be fair to denunciate unilaterally the input of our ICTs to the development of our liber-ties The Internet and ambient intelligence open the way to people “without bor-ders”, able to give to his or her speech an international and unprecedented dimension

governmental-It must be clear that the Internet liberates us from traditional normativities: “within the Internet clouds I feel free” I am able to build up my own personality by com-municating with others and to discover the knowledge generated by the whole of humanity In the same sense, we underline that ambient intelligence like the brain computer interfaces creates opportunities for dialogues with things that might be put at our service Body implants will increase our human potentialities and tomor-row bioengineering techniques will authorize an enhanced man

Nevertheless, these technical advances, even if from a certain point of view they are increasing our liberties, at the same time are creating huge risks for them and are raising fundamental questions other than the traditional ones concerning the protec-tion of our intimacy So new issues, more salient and crucial, are now entering the discussion like the question of justice as regards access to these technologies, the risk of a two-tier society, the question of democracy when we consider economico- technical broadly non transparent governmentality and the question of social justice

in relation to the consequence of profi ling applications rejecting a priori and without appeal certain categories of population The question of dignity in the Kantian sense

of the word is also to be raised since it is clear that, analysed through profi ling niques that use data collected from a large number of sources, the human is defi ni-tively not considered as an end as such but purely as a means put at the service of marketing or security logic Algorithmic governmentality operates without the pos-sibility for the human beings, who are subject to it, to challenge the reasoning behind what is proposed as a truth, precluding any discussion, criticism or debate How do we face these new challenges? Is privacy an adequate concept to answer to all these challenges and, if yes, with which meaning and how do we envisage the relationship between data protection and privacy, which are considered apparently

tech-as at letech-ast two separate human liberties by the EU Charter?

Luiz suggests the reader make a detour by scrutinizing the relationships between the Sen’s or Nussbaum’s theories of capabilities and privacy Under Sen, capabili-

ties encompass the conditions which enable the citizens to become ‘fuller social

persons, exercising their own volitions and to interact with – and infl uence- the world in which they live’ The interest of bringing closer together the concepts of

“capabilities” and “privacy” is twofold Firstly, it underlines the fact that the vidual’s mastery of his or her environment is not obvious and does not depend on his or her own volition but presupposes an active role of the state, which in a societal and economic context will enable this possibility of mastery Arendt, as noted in the

fl ourishing of our human fundamental rights and freedoms

To support his thesis, Luiz attentively analyses the case law generated by the application of Article 8 ECHR Particularly in his reading of cases like Botta vs Italy, he demonstrates the prominent place afforded to the means to freedom rather than to freedoms themselves As asserted by the German constitutional Court since

1983 in the famous census case, the right to self-development within a given etal context is an adequate criterion to defi ne the outlines of privacy requirements,

soci-considered as a tool for ‘ sustaining the uniquely human capacity for individual

refl exive self-determination and for collective deliberative decision making ing the rules of social cooperation ’ The author insists on the fact that the concept

regard-of privacy is evolutive in its concrete meaning since it will refer to different means according to the evolution of the socio-economic, technological and cultural context wherein that human capacity will have to develop itself If privacy could be limited

to the protection of home, correspondence and sensitive data in 1950, the new nologies, the globalization of our economy, the profi ling activities,… oblige us to give to privacy another dimension and to recognize new subjective rights in order to achieve our capacity for self-determination

Data protection legislation appears in that perspective as an historical answer to the risks created for our self-development by an information society and thus is directly derived from the privacy concept As asserted by the author, legislation cre-ates procedural guarantees (duty to inform, obligation to register and so on) and

subjective rights (right to object, right to access,…) in order to leave ‘ space for

individuals to choose the lives they have reason to value ’ Ambient intelligence and

the profi ling activities authorized by modern technologies oblige us to renew our legislation in different directions The fi rst one, defi nitively, is to draw our attention

to the technology itself Traditionally data protection legislations consider only the relationship between data controllers and data subjects considered as a liberal sub-ject, the relationship submitted to the DPA control From now, we have to consider the technology itself insofar as the danger resides in the software algorithms, the infrastructure and the functioning of terminals We have to take care of the potenti-alities of the technology, the design of the ICT systems, and the logic behind the algorithms Moreover, with the author we plead for a risk assessment of ICTs and for public debates about new applications and their societal impacts The second point will be to underline the crucial role of the state which has to create this space for democratic discussion and to preserve the conditions of a public sphere where every citizen might, with confi dence, express him or herself and develop his or her own personality

So, using different theoretical approaches and concepts (virtuality, capability, agency, due process, governmentality) and authors (Foucault, Sen, Rouvroy, Deleuze, Hildebrant) and combining these different sources in an original and fruit-ful reasoning at the service of the defense of human values, Luiz Costa offers the Foreword

reader inspiring refl ections aiding a better understanding of the challenges raised by ICTs today and opening new avenues towards a renewed and more adequate regulation

3rd March 2016

Acknowledgments

This book is the result of the study developed during my PhD in Law at the University of Namur

I have several reasons to thank Yves Poullet, some of these are having agreed to

be my supervisor, conveying his passion for research, being a constant, patient, zealous teacher, for suggesting the connection between my research, capabilities and the concept of privacy, for his tireless encouragement during this journey and for our friendship I thank Antoinette Rouvroy, my co-supervisor, for the opportuni-ties we had to discuss this thesis, for her patience as a teacher, for her numerous insights, inspiring vigorous thought and constant encouragement I thank Cécile de Terwangne and François Xavier-Standaert, members of the support Committee, for following my research and for their advice Cécile has encouraged me since the very beginning and throughout this work I thank Mireille Hildebrandt and Françoise Tulkens, members of the jury, for the attention they paid to this work In particular,

I am grateful for Mireille’s precious thought-provoking evaluation; I hope to have reacted to it in a way that improves this work

I thank the people that helped me either with our conversations or through their reading of the thesis, in particular Amartya Sen and Peter Burke for their generous attention and also André Silva, Katherine Albrecht and Ben Wizner I thank Joel Reindeberg for having kindly welcomed me for as a visiting researcher at Fordham University I also thank Victoria Nash and the Oxford Internet Institute for having

me on their summer doctoral programme

I thank the members of the Centre de Recherche Information, Droit et Société (CRIDS) for having welcomed me so warmly when I came to Namur A special thanks to Jean-Marc Van Gyseghem for his great company and also to Robert Queck, Laurence Lecocq, Sarah Fievet, Jacques Gerard, Virginie Marot and Quentin Houbion

I thank my teachers and colleagues at the Conservatoire Balthasar-Florence de Namur for the precious moments we had during my long stay in Namur

I thank all the members of the High Council of the Brazilian Federal Prosecution Service who agreed to give me professional leave to conclude my thesis

I thank Julia Mortimer for having edited the text and for the smoothness of our professional relationship in such an intense period

I thank Neil Olivier and Diana Nijenhuijzen for their receptiveness, patience and guidance for the preparation of the manuscript now published by Springer I also thank the blind reviewers for their inputs to this work

I thank all my family who supported me during my PhD., Rafael Costa, Ieda and Antônio, Letícia, Conceição, Ascenção, Maria Esther, Maria Lucia, Alex, Diana, Rafael and Silvana I thank also all the family Poullet and my friends

I thank Delfi na, my mother, and Ribamar, my father, for their love and for having encouraged me to study

I thank Raquel for existing and for our love, which is inspiration and life

Contents

Part I Power and Freedoms

1 Introduction 3

1.1 Setting the Stage 3

1.2 Studying Power in Order to Study Freedoms 5

1.3 Our Ariadne’s Threads 8

1.3.1 The Elimination of the Social Effects of Uncertainty and the Risks to Freedoms 8

1.3.2 Vindication of Rights 10

1.4 The Structure of this Work 12

References 12

2 A World of Ambient Intelligence 15

2.1 Ambient Intelligence, What Is It? 15

2.2 The Technology Within 18

2.2.1 Enabling Technologies 18

2.2.2 A Focus on Automated Profi ling 22

2.2.3 AmI Features 25

2.2.4 Around the Ideas of Power and Freedoms 28

2.3 Conclusions 37

References 38

3 The Power Through Technology 43

3.1 How Non-Neutral? 44

3.2 Surveillance, an Account of Power Through Technology 48

3.3 Algorithmic Governmentality 52

3.3.1 The Concept and the Means of a New Way to Govern 53

3.3.2 Virtuality 58

3.4 Conclusions 62

References 63

4 The Capability Approach 67

4.1 Presenting the Capability Approach 67

4.2 Central Concepts and Distinctions 69

4.2.1 Freedom and Means to Freedom 69

4.2.2 Sets of Beings and Doings 71

4.2.3 Evaluative Spaces: Well-Being and Agency; Freedoms and Achievements 73

4.3 Capabilities and Human Rights 75

4.4 At the Crossroads with Theories of Technology 77

4.4.1 From Access to Technology to the Conversion of Resources into Capabilities 78

4.4.2 The Impact of Technology Design on Capabilities 79

4.4.3 Capabilities and Risk 80

4.4.4 Human Development and Human Enhancement 81

4.5 Conclusions 83

4.6 An Evaluation of Part I 84

References 89

Part II Rights 5 Privacy and Human Flourishing 95

5.1 The Private and the Public 98

5.1.1 The Private, the Public and the Law 98

5.1.2 Is There Something Private in Essence? 100

5.2 Privacy, A Fundamental Human Right 104

5.2.1 The Multiple Dimensions of Privacy 104

5.2.2 Making Some Sense of Privacy: Our Starting Points 114

5.3 Crossing Languages 121

5.3.1 Vindicating Rights 121

5.3.2 Privacy as Long as It Is Free? 124

5.3.3 Privacy and Well-Being 126

5.3.4 Virtuality, Privacy and Capabilities 127

5.4 Conclusions 128

References 130

6 Data Protection Law, Processes and Freedoms 137

6.1 It’s (Almost) Nothing Personal 139

6.2 The Weak Autonomy 145

6.2.1 Out of Control 145

6.2.2 Protecting the Data Subject 147

6.3 The Techno-Legal Approach: Beyond the Good and the Evil? 152

6.4 How Transparent Should Data Processing Be? 156

6.5 From Free Flow to Just Aggregation 160

6.6 Conclusions 163

References 166

7 Conclusion 171

References 175

Appendix 177

Index 193

Contents

About the Author

Luiz Costa is a researcher specialized in information technology law He holds a masters degree from the University of Paris 1 Panthéon Sorbonne and a Ph.D from the University of Namur His main research interests are related to human rights and emerging technologies, particularly from a philosophical and legal point of view Since 2016 he has been a fellow at the Research Centre in Information, Law and Society (CRIDS) He also has 14 years’ experience as a federal prosecutor in Brazil

AC Autonomic Computing

ACHR American Convention on Human Rights

AI Artifi cial Intelligence

AmI Ambient Intelligence

ANS Autonomic Nervous System

BVerfG Das Bundesverfassungsgericht

CADHATA Ad Hoc Committee on Data Protection

CBMS Colorado Benefi ts Management System

CCSS Cloud Computing Services

CCTV Closed-Circuit Television

CITA Cellule Interdisciplinaire de Technology Assessment

CJEU Court of Justice of the European Union

CoE Council of Europe

CRID Centre de Recherche Informatique et Droit

CSD Computational Scientifi c Discovery

DHS Department of Homeland Security

DKA Data and Knowledge Analytics

DKM Data and Knowledge Management

DNA Deoxyribonucleic acid

DPD Data Protection Directive (EU Directive 95/46)

DRM Digital Rights Management

EC European Commission

ECHR European Convention on Human Rights

ECommHR European Commission of Human Rights

ECtHR European Court of Human Rights

EDPS European Data Protection Supervisor

EmNets Networked Systems of Embedded Computers

EU European Union

FAST Future Attribute Screening Technology

FIDIS Future of Identity in the Information Society

FIPPs Fair Information Practice Principles

List of Abbreviations

fMRI Functional Magnetic Resonance Imaging

GDP Gross Domestic Product

GDPR General Data Protection Regulation

GNP Gross National Product

GPS Global Positioning System

GSM Global System for Mobile Communications

HDI Human Development Index

HMI Human-Machine Interaction

ICT Information and Communication Technology

ICT4D Information and Communication Technologies for Development IETF Internet Engineering Tasking Force

IoT Internet of Things

IP Internet Protocol

IPR Intellectual Property Rights

IPv4 Internet Protocol version 4

IPv6 Internet Protocol version 6

ISP Internet Service Provider

ISTAG Information Society Technologies Advisory Group

KDD Knowledge Discovery in Databases

LBS Location Based Services

LBSN Location Based Social Networking

MIT Massachusetts Institute of Technology

ML Machine Learning

NSA National Security Agency

OECD Organization for Economic Co-operation and Development

P2P Peer-to-Peer

PARC Palo Alto Research Center

PbD Privacy by Design

PC Personal Computer

PDAs Personal Digital Assistants

PETs Privacy-Enhancing Technologies

PIA Privacy Impact Assessment

RCT Rational Choice Theory

RFID Radio-Frequency Identifi cation

SMS Short Message Service

TETs Transparency-Enhancing Tools

Ubicomp Ubiquitous Computing

UC University of California

UDHR Universal Declaration of Human Rights

UNDP United Nations Development Program

UNGA United Nations General Assembly

US United States

WiFi Wireless Fidelity

WiMAX Worldwide Interoperability for Microwave Access

Part I

Power and Freedoms

© Springer International Publishing Switzerland 2016

L Costa, Virtuality and Capabilities in a World of Ambient Intelligence, Law,

Governance and Technology Series 32, DOI 10.1007/978-3-319-39198-4_1

Introduction

1.1 Setting the Stage

“Smart” refers to those cities presented as places where there are investments in

“human and social capital”, traditional and modern communication infrastructure, wise management of natural resources and participatory governance (Caragliu et al

2009 ) 1 Songdo is one example of a “ smart city ”, projected to offer “high quality

eco-friendly living” in a world of population overload The city is part of one of the South Korea’s free economic zones, a project led by the government to offer opti-mum business conditions for foreign executives and staff Built from scratch, Songdo is meant to be a model for future cities in South Korea Once the infrastruc-ture is complete, says a government agent, it will establish a standard for the plan-ning and building of all future cities Connecting buildings and facilities through ICT s is a crucial aspect of these cities The network developed in Songdo is intended

to connect all of its components: “residences, offi ces, schools everywhere, all of the buildings” (Tanaka 2012 ) Providing an image of what an Ambient Intelligence ( A m I ) city will be like, this brief reference to Songdo is intended to show the soci-etal issues with which we are concerned Below I refer to the vision of “the city of the future” as outlined by Cisco, a multinational company that is part of the consor-tium that is building Songdo This may give us a glimpse of the kind of living its residents are expected to have

Houses will be embedded with ICT s If certain objects are touched, an array of services will be displayed This means there will be many of the communicating objects that home automation offers, like the mirror in the bathroom that is also an interface to access the internet or the fridge that informs you about the stock of basic foods inside and automatically passes a command to your preferred online grocery shop Home security and automation systems will integrate multiple devices in the

1 See also Chourabi et al., who point out that the concept of “smart cities” is emerging and defi tions focus on aspects as various as looking-forward performance, sustainability, use of ICT s and others (Chourabi et al 2012 )

2013a ) There will be no need to put the garbage out since all household waste will

be sucked directly from the kitchen to an underground network of tunnels and to waste processing centers, where it will automatically be sorted, deodorized and treated (Williamson 2013 ) Offi ce buildings will be managed in a more effi cient way: applications will turn on and off based on usage, facilities will be notifi ed of peaking energy use and security alerts will be sent when unidentifi ed people enter the building or when unidentifi ed materials are found (Cisco Systems, Inc 2013b ) Also, “Telepresence”, will connect every home, offi ce and school through video, eliminating people’s need to move from place to place Worries about traffi c jams and reaching work meetings on time will be removed since every building will be connected and there will be no need to physically attend the meeting, one will be

“always there” (Tanaka 2012 ) Public transportation will be more “intelligent” and

“fl exible” Taking advantage of GPS and wireless technologies, it will be possible

to charge “intelligent fees for road use” and communicate transport-related mation in real time Retail innovations will include customer tracking and analysis, use of “smart” walls to select clothing or other goods and personalized advertising through screens that switch context according to the profi le of the viewers 2

The “ Automated Target System-Persons ” is a system the US Department of

Homeland Security ( DHS ) uses in border control According to the Privacy Impact Assessment ( PIA) produced by the DHS :

[t]he project involves conducting research to select the specifi c sensors that will capture video images, audio recordings, cardiovascular signals, pheromones, electrodermal activ- ity, and respiratory measurements For example, one potential measurement is heart rate There are a number of technologies that a sensor can use to capture heart rate One aspect

of the research is determining which specifi c sensor technology most accurately captures the desired measurement Another aspect is reviewing the research records to determine if the measurement being captured is actually an indicator of the behavior being evaluated (i.e., did increased heart rate actually occur when the subject was intending to cause a dis- turbance) […] (United States Department of Homeland Security 2008 )

A World of Ambient Intelligence While life as projected in Songdo sounds like a synonym for convenience, purchasing power and comfort – something certainly

2 Cisco’s imagined day speaks of the possibilities of future marketing and consumption: “[i]magine you are on a bus or walking down the street Digital signage shows you personalized offers, or you receive a message from your favorite boutique just around the corner It’s a digital coupon and an invitation to a great sale that is taking place right now – just for you When you walk in the door, you’re greeted by your favorite salesperson The screen over the counter begins to display a selec- tion of items based on your recent purchases They are shown in your favorite colors If you want

to buy, use your phone and your digital coupon to pay” (Cisco Systems, Inc 2013b )

1 Introduction

arguable from an urbanistic and environmental point of view 3 – the name “Automated Target System-Persons”, betrays itself, evoking traditional representations of the power of governments of the Leviathan or Big Brother style A common trait between these examples is the way in which powerful institutions defi ne the techno-logical infrastructures that shape our societies opaquely and with unsuspected power Coupled with this “power in designing” the technological infrastructure, are the asymmetries between people and those who determine individual experience (Cohen 2012 ) These asymmetries tend to be amplifi ed in Ambient Intelligence ( A m I) , a world where objects are “smart”, environments are “sensible” and technol-ogy anticipates and satisfi es our needs and desires, a world where “[t]echnology will become invisible, embedded in our natural surroundings, present whenever we need it, enabled by simple and effortless interactions, attuned to all our senses, adaptive to users and context and autonomously acting” (Lindwer et al 2003 ) Rising with the assemblage of already existing and developing multiple termi-nals, infrastructure and services, A m I tends to dismiss the very need to act inten-tionally; after all, it is about technologies that are precisely conceived to avoid awareness (systems are non-noticeable) and conscious actions to trigger systems Songdo and the above-mentioned system of border control are examples of A m I applications If A m I systems can contribute to human development and enhance-ment – for instance, liberating us from repetitive tasks or opening new possibilities for action – the promise of convenience, better life and a terrorist-free world will nevertheless be received with a fair dose of skepticism If freedoms matter – we may think about freedom of choice or freedom of action – one may reasonably be con-cerned about “environments” whose structures are determined by private and public institutions and that have a high potential to determine individual “experience”

1.2 Studying Power in Order to Study Freedoms

Subject to Power Faced with structures over which he or she has very little infl ence, the individual is put in a vulnerable position I illustrate this point with the fact that the individual – or the “data subject” as described in data protection legisla-tion – is meant to legitimate, through individual will, huge schemes of data process-ing of which he or she has little, if any, knowledge Making any sense of the fair concerns about how technology infl uences our freedoms is not an easy task In an attempt to have an idea of “what is going on”, my conceptual starting point is

u-“power”, here taken in the Foucauldian sense of “action upon action”:

what defi nes a relationship of power is that it is a mode of action that does not act directly and immediately on others Instead, it acts upon their actions: an action upon an action, on

3 Living in urban areas seems to be hardly reduced by such an idealistic projection of a city; besides, if Songdo’s lifestyle is projected to a world scale, issues such as the depletion of natural resources will certainly emerge

possible or actual future or present actions A relationship of violence acts upon a body or upon things; it forces, it bends, it breaks, it destroys, or it closes off all possibilities […] A power relationship, on the other hand, can only be articulated on the basis of two elements that are indispensable if it is really to be a power relationship: that ‘the other’ (the one over whom power is exercised) is recognized and maintained to the very end as a subject who acts; and that, faced with a relationship of power, a whole fi eld of responses, reactions, results, and possible inventions may open up (Foucault 1997 , 340)

Power Through Technology The “quality of being a subject who acts” and “the possibility of a whole fi eld of responses” on the part of the one over whom power is exercised – to mention just two elements of the Foucauldian defi nition – are expected

to face severe storms with A m I technologies Without anticipating the issues of this study, I, note, nevertheless, that my interest in power is instrumental to an investiga-tion of freedoms In particular, I am interested in the protection of freedoms in rela-tion to “power through technology”, about which I make initial points below Aware that the dimensions of power – economic, political and technological – operate in ways that are not always clearly distinguishable, I pay particular attention

to how technology infl uences action A world where desires are “anticipated” and objects and environments are “smart” and sensible to our needs depends on techni-cal capabilities that are, at the same time, highly complex and decisive of peoples actions, as illustrated by automated profi ling Observation and prediction being essential in a world of A m I, automated profi ling is the mechanism that makes it possible for systems to “read, interpret and anticipate” human behavior Built on the processing of huge amounts of data, the knowledge underlying automated profi ling

is, we will see, less objective than it may appear

With power through technology I refer to technology as a technique to direct the conduct of individuals or groups – action upon action, in the Foucauldian sense Power through technology is a subject that has received a good deal of attention from philosophers of technology, concerned with issues such as technological neu-trality, agency and risk 4 The role of technological structures in infl uencing human action is not new, as indicated by a part of privacy theory – one of the main research

fi elds in this work As far as A m I is concerned, relatively few studies have paid attention to privacy related issues, and most of those are in the philosophy fi eld The relative under-representation of legal theory in “ A m I and privacy” issues has, to a certain extent been compensated for in recent years by a growing attention to auto-mated profi ling and its regulation A fi rst gap in research is worth noting here Privacy legal theory in general and legal philosophy works about A m I pay little

attention to the investigation of technology per se , in other words to the exploration

of the specifi c rationality underlying its functioning To this purpose, the explanation

4 As is the case with the “critical theory of ICT ” which, as Zheng and Stahl point out, sees agency

as infl uenced by societal structures that may limit personal freedoms, indicating that technology may serve as a hegemonic means while rendering ideologies invisible and it is not necessarily a synonym for desirable progress (Zheng and Stahl 2012 ) See Franssen et al for a broad review of the ethics of technology literature on power issues (Franssen et al 2013 )

1 Introduction

frequently evoked is the surveillance account, which, we will see, provides a fair but insuffi cient explanation of such rationality

Algorithmic Governmentality I intend to explore this gap by returning to the

algorithmic governmentality hypothesis, which describes the form of power based

on profi ling algorithms and on the structuring of the possible fi elds where als may act (Rouvroy 2010 ) Algorithmic governmentality takes fully into account how knowledge built on machine learning is used for prediction and preemption 5 Pointing out the indistinction or the fusion between the knowledge of algorithms and power, algorithmic governmentality also makes evident that the potential dimension of human beings 6 must be preserved in order to open the possibility of free relations with technologies

Freedoms and the Capability Approach Power through technology affects the multiple experiences of human freedom Here, I take freedoms in a broad sense, an attitude refl ecting attention turned not only to the formal declaration of freedoms –

of which the rights to privacy and data protection are examples – but also to their effective realization – for instance taking into account how technologies impact on well-being and what effective choices are open to individuals in an A m I environ-ment This language of freedoms, still too loosely specifi ed at this point, underlies

the capability approach , a theoretical framework pioneered by Sen and Nussbaum

and whose two core normative claims are that “freedom to achieve well-being is of primary moral importance” and that “freedom to achieve well being is to be under-stood in terms of people’s capabilities” (Robeyns 2011 ) In the capability approach, freedoms are understood in the sense of sets of “beings and doings” or the “real opportunities” people have to “choose the lives they have reason to value” My interest in the capability approach was initially sparked by its focus on assessing real opportunities for freedom people have

Introducing the capability approach in legal studies, as I will do here, is not new

In particular, in the domain of economic and social rights, legal scholars have cussed, for instance, how to capture capabilities in a body of constitutional rights 7 Closer to our research fi eld, Cohen has explored the approach to assess the norma-tive foundations of information law and policy and to ground a theory of the net-worked self (Cohen 2012 ) I intend to explore, inter alia, the possible connections between the capability approach, privacy and data protection The object of public debate for a long time, privacy is at the center of numerous controversies from the understanding of its fundamentals to the role it is meant to play Frequently attached

dis-to the protection of personal audis-tonomy – undersdis-tood as the capacity of someone dis-to govern oneself – the relevance of privacy spreads over different domains of life,

5 See in particular Sect 3.1 for the meaning of preemption, which for the moment I generally retain

as a kind of qualifi ed anticipation

6 This is the core of the concept of virtuality, which will be described in Sect 3.3.2

7 See in this sense the interventions of Buss and Ellickson in the Conference ‘ Creating Capabilities ’ ( Creating Capabilities: Sources and Consequences for Law and Social Policy 2010 )

from its very constituency (one’s own body for instance) to the development of personality and also to the mastery or control of information, assumed to be of importance to protect freedoms in our information societies

Research Questions With this work I intend to fi ll a gap in legal studies concerned with A m I technologies I will seek to understand the technological outline of A m I

as well as the algorithmic governmentality hypothesis and the capability approach, grounding a theoretical framework from which I will explore legal problems related

to privacy and data protection

My fi rst research question is: “Is a theoretical exploration of the algorithmic governmentality combined with the capability approach useful for a better under-standing of power and freedoms in a world of A m I ?” I will deal with two sub- questions in Chap 3 , namely: “Does algorithmic governmentality provide an advantageous explanation for the issue of power through technology?” and “Is the philosophical concept of virtuality useful for the promotion of freedoms?” In Chap

4 I will deal with two other sub-questions: “Does the capability approach provide added value to the theory of human rights and/or contribute to their promotion?” and “Is a technological version of the capability approach envisageable?”

My second research question is: “Are these theories useful for a better hension of privacy and data protection concepts and the evolution of their regula-tions?” In Chap 5 I deal with the sub-question “Is there any possible connection between privacy theory and the capability approach and between privacy and virtu-ality?” and in Chap 6 with “What challenges does A m I bring to data protection law and what does the theoretical approach built here bring to the promotion of human rights related to data processing?”

In this process I will of course deal with the theoretical framework and debates concerning algorithmic governmentality and the capability approach – as well as with the traditional debates to which privacy and data protection legal theory have been linked

1.3 Our Ariadne’s Threads

1.3.1 The Elimination of the Social Effects of Uncertainty

and the Risks to Freedoms

Living in a World of Uncertainty A world of A m I is highly informed by a nale of anticipation and pre-determination of futures, a sort of medicine to cure the diseases born within uncertainty and unpredictability This rationale is perhaps a sign of a paradoxical relationship between our ignorance and the will to master natural and artifi cial processes In other words, says Anker, we live “in a constant reminder of our limits, our inability to forecast and absolutely determine our pro-jected future And yet, we continue to press ahead with our planning, as if we were

ratio-1 Introduction

not staring directly into a future of incessant indeterminacy” (Anker 2006 , 44) One

of the threads that will guide this study is precisely our relationship with regard to uncertainty; or, more particularly, the risks that approval of the elimination of the social effects of uncertainty poses to our freedoms Aiming at a colorful narrative of this thread I refer to two distantly related analogies around the idea of error The fi rst is related to cultural transmission, where Dawkins mentions the fi ndings

of Jenkins about bird singing 8 :

[o]n the island where he [Jenkins] worked there was a total repertoire of about nine distinct songs Any given male sang only one or a few of these songs The males could be classifi ed into dialect groups For example, one group of eight males with neighbouring territories sang a particular song called the CC song Other dialect groups sang different songs Sometimes the members of a dialect group shared more than one distinct song By compar- ing the songs of fathers and sons, Jenkins showed that song patterns were not inherited genetically Each young male was likely to adopt songs from his territorial neighbours by imitation, in an analogous way to human language During most of the time Jenkins was there, there was a fi xed number of songs on the island, a kind of “song pool” from which each young male drew his own small repertoire But occasionally Jenkins was privileged to witness the “invention” of a new song, which occurred by a mistake in the imitation of an old one He writes: “New song forms have been shown to arise variously by change of pitch

of a note, repetition of a note, the elision of notes and the combination of parts of other existing song” (Dawkins 1989 , 189–190)

Error in pathology is the second one, and I highlight below the warning of Canguilhem concerning the dangers of developing medical counterparts to heredi-tary errors:

It’s possible to neutralize the effects of an error of metabolism by constantly furnishing the organism with the reaction product which is indispensable to the exercise of that function […] But to compensate an organism’s defi ciencies for life only perpetuates a solution of distress The real solution to heresy is extirpation Consequently why not dream of hunting for heterodox genes, of a genetic inquisi-tion? And while waiting, why not deprive suspect sires of the liberty of sowing broadcast? We know that these dreams are not only dreams for some biologists of very different philosophical persuasion, if we may call it that But in dreaming these dreams, we enter another world, bordering on the bravest of Aldous Huxley’s from which sick individuals, their particular diseases and their doctors have been elimi-nated The life of a natural population is portrayed as a lotto sack and the functionar-ies delegated by the life sciences have the task of verifying the regularity of its numbers before the players are allowed to draw them from the sack to fi ll their cards At the beginning of this dream we have the generous ambition to spare inno-cent and impotent living beings the atrocious burden of producing errors of life At the end there are the gene police, clad in the geneticists’ science […] (Canguilhem

1991 , 280–281)

On Totalization Non-predicted errors, deviations and miscalculations seems to be

a common trait in the trajectory of birds and human beings While the cultural

8 I specially thank André Silva for calling my attention to the example

transmission amongst the birds example talks about the creative force of error in producing new songs, the eugenics example points to dangerous paths that the elim-ination of heterodox genes may take What I highlight from these analogies is that

in the fi rst example the non-predicted, non-anticipated error has a positive value, i.e., making a “mistake” was a good thing as it enlarged a song repertoire The sec-ond is that the elimination of uncertainty is the very atrocity it claims to eliminate and history has given quite sinister examples on how far such elimination may go The violence within the last example is obvious and I have no intention of putting this example on an equal basis with the technologies with which we will deal here Nevertheless, and if is there any possible common core amongst bird singing, eugenics and A m I technologies, it is the idea of “totalization” which, as says Anker,

is about attempting “to close down the future and give nothing other than what is and what is already known It gives us a world of calculation and pre-existing knowledge in the here and now, but it cannot give us a future which holds the poten-tiality of an-other, a some-thing other, a thought not yet thought or determined by the present conditions” (Anker 2006 , 53)

A sign of such totalization is the design of A m I technologies, which are – we will see soon in detail – ubiquitous, i.e., spread everywhere and meant to connect every-thing in networks directed by an overarching, totalizing, logic 9 Being unable to eliminate uncertainty itself, a world of A m I is about avoiding the undesirable social effects of such uncertainty We will see throughout this study how such logic is put

in motion; for now I emphasize that preserving the unpredictable and the potential

is important not only in a philosophical sense but also in a practical one, our doms depend on such preservation; again according to Anker

[d]emocracy […] needs a future which never absolutely arrives […] The infi nite arrival, the

“to come”, does not entail a passivity toward a working and striving toward a democratic state, it only recognizes that the notion of absolute arrival, an absolute here and now of a democracy without a future always to come, contains the seeds of totalized thinking Democracy needs the “to come” of the future or it is not democracy (Anker 2006 , 53–54)

1.3.2 Vindication of Rights

Realizations and Accomplishments Sen distinguishes theories of justice that focus on the idea of a just society and those that focus on social realizations, both having come out in the Enlightenment period While the former is concerned with the question “what would be perfectly just institutions?” the latter is concerned with the question “how should justice be advanced?” Hobbes and later Locke, Rousseau

9 As pointed out by Fox and Kemp “[…] we are seeing that we now have the potential to think of space as being organized in a computational network Objects can have both the fundamental logic and hardware to allow them to be extremely good at executing specifi c tasks they were intended to

do while simultaneously networking into a collective whole that can be controlled by an ing logic The idea of ubiquitous computation is about embedding hardware and software, infor- mation processors and coded intelligence in all aspects of our lives” (Fox and Kemp 2009 , 62)

overarch-1 Introduction

and Kant are representatives of the fi rst group of theories and Smith and later Condorcet, Wollstonecraft, Bentham, Marx, Mill and others of the second Sen refers to the fi rst group of theories as “transcendental institutionalism”, as they pose the question of justice in an abstract, idealistic manner Taking a different path, he embraces the second tradition, which focuses on actual realizations and accom-plishments rather than only on institutions and rules, on realizations rather than on arrangements This choice is explained in part by the argument that justice cannot

be indifferent to lives that people can live:

the importance of human lives, experiences and realizations cannot be supplanted by mation about institutions that exist and the rules that operate Institutions and rules are, of course, very important in infl uencing what happens, and they are part and parcel of the actual world as well, but the realized actuality goes well beyond the organizational picture, and includes the lives that people manage – or do not manage – to live (Sen 2009 )

Realization and the Advancement of Human Rights is the second thread of this

work In order to clarify this thread I invoke Sen’s response to the question “What

are human rights?” In his attack on the French Declaration of Human and Civil

Rights of 1789 , Bentham opposed the “real right”, the child of the law, to the inary right”, the child of imaginary laws, of the law of nature (Bentham 1843 ) Bentham argued that a claim, to be considered as a right, must have legal force In opposition to this understanding Sen sets up that of Wollstonecraft, for whom human rights demanded ethical recognition of “the acknowledged rights of every-one, in the form of respecting freedoms and corresponding obligations” (Sen 2009 ) Far from the strict sense where freedom is synonymous with non-interference by third parties, for Wollstonecraft, “it is unsustainable to have a defence of the free-dom of human beings that separates some people whose liberties matter from others not to be included in that favoured category” 10 ; “vindication of rights” is the idea of the obligation of effective power to advance the freedoms of all According to Sen [p]roclamations of human rights, even though stated in the form of recognizing the exis- tence of things that are called human rights, are really strong ethical pronouncements as to what should be done They demand acknowledgement of imperatives and indicate that something needs to be done for the realization of these recognized freedoms that are identi-

“imag-fi ed through these rights (Sen 2009 )

As a legal researcher myself, I certainly to not deny the legal aspects of human rights, and I will deal particularly with those related to the right of privacy However, while acknowledging such a dimension I retain the view that up to now human rights are not “legal rights” in the sense of Bentham, meaning that they are not syn-onymous with subjective rights, nor are they are discussable only where they are realized 11 Besides their legal relevance, human rights are also about making not yet

10 Wollstonecraft’s writing, notes Sen, was aimed not only at the iniquities suffered by women “but also at the treatment of other deprived groups of people, for example slaves in the United States and elsewhere” (Sen 2009 )

11 I will later evaluate Deleuzian concepts of the real, the virtual and the possible as well as virtuality as defi ned by Rouvroy – see in particular Sect 3.3.2 For the moment I retain the idea of

feasible rights feasible, as pointed out by Wollstonecraft and later Sen The idea of the vindication of rights inspires this study

1.4 The Structure of this Work

In Chap 2 I will review how technical literature has described the assemblage of technologies that make possible a world of A m I , paying particular attention to its technical features and more specifi cally to machine learning The technical outline will be followed by a general discussion of societal issues, and more precisely of the relevance of power through technology and freedoms, subjects that will be devel-oped in the two following chapters In Chap 3 I will review several signs the litera-ture highlights as the traces of the non-neutrality of technology Taking these traces

as presuppositions of the study, I will then review the “surveillance account” of power through technology followed by “algorithmic govermentality”, the account

on which I will pave the route of the study Chap 4 will be dedicated to a general presentation of the capability approach, to the exploration of its connection to human rights and to a review of several outcomes of recent literature at the cross-

roads with technology studies With “ An Evaluation of Part I ” I consider the

previ-ous chapters and establish the grounds for reading the legal problems of privacy and data protection in a world of A m I as oriented in the second part Chap 5 outlines a general review of the state of the art of privacy, having particularly in mind the developments of legal theory and the case law of Article 8 of the European Convention of Human Rights ( ECHR ) This overview will be followed by an explo-ration of connections between privacy and the theoretical framework established in the fi rst part of this work Chap 6 will be dedicated to an analysis of how A m I chal-lenges data protection legislation as well as to speculative thoughts on how data protection law may better the advancement of freedoms

References

Anker, M 2006 ‘The Ethics of Uncertainty: Aporetic Openings’ Switzerland: European Graduate School

Bentham, J 1843 ‘Anarchical Fallacies; Being an Examination of the Declaration of Rights Issued

during the French Revolution’ In The Works of Jeremy Bentham Vol II

Canguilhem, G 1991 The Normal and the Pathological New York: Zone Books

Caragliu, A., C Del Bo, and P Nijkamp 2009 ‘Smart Cities in Europe’ Journal of Urban

Technology 18 (2): 65–82

Chourabi, H., T Nam, S Walker, J.R Gil-Garcia, S Mellouli, K Nahon, T.A Pardo, and H J

Scholl 2012 ‘Understanding Smart Cities: An Integrative Framework’ In 2012 45th Hawaii

International Conference on System Science ( HICSS ) , 2289–97

“realization” of rights as making rights effective or operational

1 Introduction

Cisco Systems, Inc 2013a ‘Cisco Builds All-Digital Wireless-Based Home Security and

Automation Control Panel for AT&T ’ Cisco’s The Network January 7 http://newsroom.cisco com/release/1122064

Cisco Systems, Inc 2013b ‘Cisco Visualization | The City of The Future’ September 4 http:// share.cisco.com/TheCityOfTheFuture/the-city-of-the-future.html

Cohen, J E 2012 Confi guring the Networked Self New Haven: Yale University Press

Creating Capabilities: Sources and Consequences for Law and Social Policy 2010 Chicago Dawkins, R 1989 The Selfi sh Gene Oxford; New York: Oxford University Press

Foucault, M 1997 ‘The Subject and Power’ In Essential Works of Foucault: 1954–1984 , edited

by P Rabinow New York: The New Press

Fox, M., and M Kemp 2009 Interactive Architecture New York: Princeton Architectural Press Franssen, M., G.-J Lokhorst, and I Van de Poel 2013 ‘Philosophy of Technology’ In The

Stanford Encyclopedia of Philosophy , edited by E N Zalta http://plato.stanford.edu/archives/ win2013/entriesechnology/

Lindwer, M., D Marculescu, T Basten, R Zimmermann, R Marculescu, S Jung, and E Cantatore

2003 ‘Ambient Intelligence Visions and Achievements: Linking Abstract Ideas to Real-World

Concepts’ In Design, Automation & Test in Europe Conference & Exhibition Vol 1 Los

Alamitos, California: IEEE Computer Society

Robeyns, I 2011 ‘The Capability Approach’ In The Stanford Encyclopedia of Philosophy , edited

by E N Zalta http://plato.stanford.edu/archives/sum2011/entries/capability-approach/ Rouvroy, A 2010 ‘Détecter et prévenir : les symptômes technologiques d’une nouvelle manière

de gouverner’ In L’état des droits de l’homme en Belgique : rapport 2009–2010 , 9–16

Bruxelles: Aden

Sen, A 2009 The Idea of Justice Cambridge: Belknap Press of Harvard Univ Press

Tanaka, W 2012 ‘Cities of the Future: Songdo, South Korea – A New Approach – The Network:

Cisco’s Technology News Site’ Cisco’s The Network January 4 http://newsroom.cisco.com/ feature-content?type=webcontent&articleId=630153

United States Department of Homeland Security 2008 ‘Privacy Impact Assessment for the Future Attribute Screening Technology ( FAST ) Project’ United States Department of Homeland Security

Williamson, L 2013 ‘Has “Smart” Songdo Been a Success?’ BBC , September 2, sec Technology

http://www.bbc.co.uk/news/technology-23757738

Zheng, Y., and B C Stahl 2012 ‘Evaluating Emerging ICT s: A Critical Capability Approach of

Technology’ In The Capability Approach, Technology and Design , edited by I Oosterlaken

and J Van den Hoven, 57–76 Springer

© Springer International Publishing Switzerland 2016

L Costa, Virtuality and Capabilities in a World of Ambient Intelligence, Law,

Governance and Technology Series 32, DOI 10.1007/978-3-319-39198-4_2

Chapter 2

A World of Ambient Intelligence

Ambient Intelligence (AmI) describes a world where objects are “smart”, ments are “sensible” and technology anticipates and satisfi es our needs and desires AmI and similar technological visions have received a good deal of attention from researchers in various fi elds In this chapter I briefl y review the profusion of techni-cal literature that names and describes the ensemble of technologies that make this world possible The focus on AmI technical features and machine learning will be followed by a general discussion of societal issues, from which I will highlight two:

environ-“power through technology” and “freedoms” The purpose of this chapter is to introduce such issues, which will be further developed in the next two chapters

2.1 Ambient Intelligence, What Is It?

Buzz Words The idea of spreading computing everywhere was behind the work of researchers during the 1990s, in company laboratories such as Nokia, Philips and Samsung and university laboratories such as MIT and UC Berkeley (Greenfi eld

2006 , 12) While designing prognostics for the future of information and cation technologies (ICTs), companies and research groups have given different names to visions, whose meaning is not always clear for two apparent reasons First, defi ning speculative technology visions demands efforts of abstraction, which, by defi nition, make understanding diffi cult Second, it is common amongst technology but also law and philosophy scholars to refer to these visions in an interchangeable manner In the following paragraphs I will review the discussion of these techno-

communi-logical visions in the literature, my purpose being to justify the choice of Ambient

Intelligence (AmI) The fi rst necessary clarifi cation is that when referring to visions

I have in mind technologies and applications that are both present and future

Weiser coined the term Ubiquitous Computing (Ubicomp) to refer to the use of

computers everywhere in a natural and unobtrusive way Back in 1991 he identifi ed

location and scale as two challenges of central importance in such Ubicomp visions

At that time computers “did not know where they were” and terminal design was PC-centered Smartphones, personal digital assistants (PDAs), tablets and other gadgets clearly overcame these challenges in subsequent years and the work of Weiser’s team at the Palo Alto Research Center (PARC) evolved accordingly, con-tributing to fi elds of computer science that speculated about “a physical world richly and invisibly interwoven with sensors, actuators, displays, and computational ele-ments, embedded seamlessly in the everyday objects of our lives and connected through a continuous network” (Weiser, Gold, and Brown 1999 , 693–696) The most profound technologies, says Weiser, are those that disappear, meaning that once people learn how something works, they cease to be aware of it – disappear-ance for Weiser, we see, is a synonym for unawareness rather than vanishing The assumption gives rise to “a new way of thinking about computers in the world, one that takes into account the natural human environment and allows the computers themselves to vanish into the background […]” (Weiser 1991 )

The term Internet of Things (IoT) describes a vision where the Internet embraces

everyday objects The expression was coined in 1999 by Kevin Ashton, co-founder

of the Auto-ID Center at the Massachusetts Institute of Technology (MIT), which pioneered the development of an Internet-like infrastructure for tracking goods globally through the use of radio-frequency identifi cation (RFID) tags (Ashton

2009 ) Use of the expression has increased since then, embraced for instance in Europe by the European Commission 1 (EC), who emphasized the role of RFID technology; at the same time it referred to a broader comprehension of IoT as a step

in the evolution of the Internet In the words of the EC, IoT is a new paradigm for the development of the Internet, one major step in this development being

to progressively evolve from a network of interconnected computers to a network of connected objects, from books to cars, from electrical appliances to food, and thus create an

inter-“Internet of things” (IoT) These objects will sometimes have their own Internet Protocol addresses, be embedded in complex systems and use sensors to obtain information from their environment (e.g food products that record the temperature along the supply chain) and/or use actuators to interact with it (e.g air conditioning valves that react to the presence

of people) (EU Communication on the Internet of Things)

Autonomic Computing (AC) is a term coined by IBM, which refers to a new

para-digm in computing The assumption is that the rationale of ICT infrastructure opment threatens to undermine the benefi ts ICTs aim to provide The causes would

devel-be the very complexity of such infrastructure and the fact that there will not devel-be enough skilled people to manage it With the evolution of computing there is a call for decision-making and response, the speediness of which surpasses human capac-ities Given complexity as the problem, IBM proposes more complexity as the solu-tion: computer systems must be able to run themselves, their main attributes being

to self-confi gure, self-heal when confronted by malfunction, self-optimize work and self-protect against attacks Horn enumerates the eight key elements of AC in the following terms:

1 Also known as the Commission of the European Communities

1 To be autonomic, a computing system needs to “know itself”—and comprise components that also possess a system identity […] 2 An autonomic computing system must confi gure and reconfi gure itself under varying and unpredictable conditions […] 3 An autonomic computing system […] always looks for ways to optimize its workings […] 4 An auto- nomic computing system must perform something akin to healing – it must be able to recover from routine and extraordinary events that might cause some of its parts to malfunc- tion […] 5 […] an autonomic computing system must be an expert in self-protection […]

6 An autonomic computing system knows its environment and the context surrounding its activity, and acts accordingly […] 7 An autonomic computing system cannot exist in a hermetic environment […] 8 […] an autonomic computing system will anticipate the opti- mized resources needed while keeping its complexity hidden (Horn 2001 )

Ambient Intelligence (AmI) raised as a concept to defi ne unobtrusive

technolo-gies that are capable of adapting to human needs; AmI is a vision where technology becomes invisible, embedded, ever-present, enabled by effortless interactions, adjusted to our senses, adaptive and autonomous (Lindwer et al 2003 ) The Information Society Technologies Advisory Group (ISTAG) of the European Commission (EC) and Philips give the two most important accounts of AmI The ISTAG, which was set up to advise the EC on its overall strategy on information society technologies, hypothesizes that in AmI environments human beings will be surrounded by intelligent interfaces supported by computing and networking technology that is embedded in everyday objects such as furniture, clothes, vehicles, roads and smart materials […] AmI implies a seamless environment of computing, advanced networking technology and specifi c interfaces This environment should be aware of the specifi c char- acteristics of human presence and personalities; adapt to the needs of users; be capable of responding intelligently to spoken or gestured indications of desire; and even result in sys- tems that are capable of engaging in intelligent dialogue Ambient Intelligence should also

be unobtrusive – interaction should be relaxing and enjoyable for the citizen, and not involve a steep learning curve (ISTAG 2005 , 47)

To Philips, a Dutch technology company, AmI introduces

a common view on the development of future electronic products and services […] Lighting, sound, vision, domestic appliances and personal healthcare products and services all communicate and cooperate with one another to improve the total user experience […] the vision of Ambient Intelligence has brought a common and tangible focus to a variety of endeavors, encouraging many in our organization to see how their activities fi t into the overall concept of an environment that is sensitive and responsive to the presence of people and sympathetic to their needs (Harwig 2006 )

While Philips’ defi nition is derived from a consumption perspective, ISTAG’s sounds broader, from a perspective that was later confi rmed by the EC, which endorses AmI as a paradigm for technological development and announces the age

of Ambient Intelligence The way the EC departs from the idea of the AmI home applications to others is representative of a shift from a consumer, private perspec-tive to a general one, involving applications vis-à-vis persons not only as consumers but also as citizens, workers, commuters and so on:

[…] devices throughout the house can be in constant contact with each other, making the AmI home responsive to all its inhabitants’ needs thanks to a variety of services which can

be easily installed and personalised to the user’s needs […] Similar scenarios can be aged for other environments such as the car, the offi ce, leisure and cultural settings, public 2.1 Ambient Intelligence, What Is It?

envis-spaces and various general interest services A key feature is the ability for seamless ment between these spaces: people on the move will become networks on the move as the devices they carry network together and connect with the different networks around them (European Commission 2004 )

Why Ambient Intelligence Finally? A brief comparison between the presented visions 2 points to overlaps Weiser’s Ubicomp anticipated the diffusion of network computing that would appear with the diffusion of the Internet also being the basis

of the vision of pervasive computing, a technological paradigm where computers assume multiple forms and are spread throughout the environment (Ahonen et al

2008 , 2) Built upon the concept of Ubicomp, AmI is more specifi c concerning the role of environments, which are fully embedded with ICTs and capable of sensing, adapting to and responding to the presence of human beings and supporting them in daily life (Surie 2012 ) AmI embraces both IoT – which is bound by the idea of con-necting things through the Internet – and AC – grounded on the principle of self- suffi cient computing systems AmI relies on technologies that also exist in other domains such as Ubicomp, IoT and AC In order to maintain the continuity of this study, I use AmI for its broader scope, covering potentialities of technology that are sparsely treated in other visions 3 I also use AmI for a practical reason: since AmI narrative is especially widespread in Europe and this study is particularly concerned with the European legal perspective on emerging technologies, referring to AmI is inevitable Nevertheless, I will sometimes refer to other visions when emphasizing technological features with which they specifi cally deal

2.2 The Technology Within

From a technological point-of-view, AmI can be seen as a property of multiple minals, infrastructure and services My purpose at this point is to mention technolo-gies that set the tone of the AmI I roughly follow the fi ndings of Ahonen et al.,

ter-which categorize the enabling technologies of AmI in ubiquitous computing

tech-nologies , ubiquitous communications , user-friendly interfaces , embedded gence and sensors and actuators (Ahonen et al 2008 )

intelli-2 Other technology descriptions are similar to those mentioned above such as Ahonen’s “ubiquitous networked society” (Ahonen et al 2008 ) and Greenfi eld’s “everyware” (Greenfi eld 2006 ) Each of these visions has a particular approach to ICT and focuses on different aspects of a technology continuum that involves networks, communications and terminals While acknowledging the value

of these visions, I have no wish to further deepen distinctions between them

3 For instance, the EC, while describing AmI systems, refer to attributes that are embraced by autonomic computing, namely: systems that know themselves, are dynamic, self-optimizing, resil- ient and so on (European Commission 2004 )

Ubiquitous computing technologies are those that will make possible the spread of computers everywhere, as RFID 4 illustrates RFID is an identifi cation system, based on electromagnetic waves, which makes possible contactless identi-

fi cation, processes data through tags and readers 5 and connects to a network structure Nowadays 6 RFID is used in various domains such as the automotive industry (as in hands free car kits), building and area access control systems, chip cards (as credit and transport cards) and identity documents (such as passports and identity cards) Its uses in the industry and services sectors are numerous 7 AmI technologies are ubiquitous as they can literally be everywhere

Ubiquitous communications refers to the electronic communications

infra-structure necessary to connect everything AmI certainly profi ts from existing work infrastructures such as cloud computing 8 and peer-to-peer (P2P), 9 as well as wireless technologies – such as global positioning system (GPS), global system for mobile communications (GSM) and technological protocols such as bluetooth, wireless fi delity (WiFi) and wireless interoperability for microwave access (WiMAX) Though wireless communications play an important role in AmI, con-necting everything requires a high level of interoperability “between all kinds of

net-4 The EC defi nes RFID technology as “the use of electromagnetic radiating waves or reactive fi eld coupling in the radio frequency portion of the spectrum to communicate to or from a tag through a variety of modulation and encoding schemes to uniquely read the identity of a radio frequency tag

or other data stored on it” (Article 3, a, of the Commission Recommendation of 12.5.2009 on the Implementation of Privacy and Data Protection Principles in Applications Supported by Radio- Frequency Identifi cation, hereafter the “RFID Recommendation”)

5 A RFID tag or chip is a device that produces a radio signal or responds to and modulates a carrier signal received from a reader A reader is a device that captures and identifi es electromagnetic waves

6 RFID initial uses were military as in the case of the Friend or Foe application, created by the British Royal Air Force in order to identify enemy airplanes during World War II (Avoine 2009 , 17)

7 I would also mention applications such as supply chain management, identifi cation of packages, car hands-free ignition, collection systems in toll roads, control of access to buildings, baggage handling control in airports, contactless credit cards, identity cards, passports and medical records

8 Cloud computing “refers to applications delivered as services over the Internet as well as to the actual Cloud infrastructure—namely, the hardware and systems software in data centers that pro- vide these services [moving] computing and data away from desktop and portable PCs into large data centers” “The main technical underpinnings of cloud computing infrastructures and ser- vices”, say Dikaiakos et al “include virtualization, service-oriented software, grid computing tech- nologies, management of large facilities, and power effi ciency” (Dikaiakos et al 2009 , 10) Cloud computing delivers services through the Internet from resource clouds where information is pre- served This means that computing resources are used with great effi ciency and are easily acces- sible Cloud computing is therefore a powerful instrument to enable connectivity in an AmI world without regard to how or where data is stored (Veikko et al 2010 , 78; Rader et al 2011 , 47)

9 P2P is a network architecture based on the partition of tasks between peers Compared to server architectures, P2P has more scalability – the ability to handle growing amounts of tasks – and robustness – maintaining the availability of the system despite mal or non-functioning of one

client-or mclient-ore peers; also, P2P has the advantage of distributing the costs of the netwclient-ork among peers Because P2P architecture is dynamic and distributed, it can be very effective in the deployment of AmI (Gasson and Warwick 2007 , 42)

2.2 The Technology Within

short-range and long-range wireless and wired networks (body area networks, sonal area networks, virtual home environment, ad hoc, cellular, sensor, satellite networks, etc.) and their convergence into all-IP all over the world” (Ahonen et al

per-2008 , 11)

“User-friendly” Interfaces involve a wide range of new ways of interacting

with ICTs more “naturally” While traditional computing uses interfaces like a mouse, keyboard and monitors, AmI interfaces employ the sensing of movement, scent and sound for instance (Alcañiz and Rey 2005 , 4) Text but also speech, ges-tures and even emotions can be used by systems in order to enhance preciseness of communication; our surroundings and even our bodies become the interface for the delivery of its services Computers take the form of a wide array of objects – such

as glasses, watches or armbands – and are connected to other devices, which may also function as personal assistants (Michael 2013 ) Interaction is meant to occur in

a more seamless way and AmI systems will be able, for instance, to understand and respond to an action based on the recognition of emotions by voice, image and physiological measurements (Ahonen et al 2008 , 15) Other examples are brain computer interfaces (BCIs) 10 – which make it possible to control mechanical mem-bers, walk and communicate through computers using mental activity alone – as well as cochlea implants, electrical stimulation of the central nervous system and retinal implants – where data is internalized from the device to the person and exter-nalized from the person to the device These bi-directional systems increase human senses, memory and thinking (Gasson and Warwick 2007 , 22; Nicolelis 2011 )

Sensors and actuators The idea of responsive environments is that their

trigger-ing must not depend on special attention from human betrigger-ings In order for this to happen, AmI will comprise sensors that capture information related, for instance, to positioning, physiology and health condition (Ahonen et al 2008 , 16) This trend in sensing is already visible today – as with smartphones 11 – and is leading to both quantitative and qualitative improvements Quantitative improvements mean, for instance, the amplifi cation of location-based services such as those which use GPS and RFID technologies, while qualitative improvements mean that more elements will be translated into inputs, such as environmental temperature or the presence of

or signals from another person Moreover, AmI will materialize its outputs through actuators, i.e., technologies accomplish the outcomes of AmI for instance “switch-ing on and off diverse home and offi ce appliances, health maintenance systems, transportation systems (e.g., driving safety) and access control systems” (Ahonen

et al 2008 , 16) Recent trends in biometrics 12 are illustrative of the new sensing

10 BCIs are pathways between the brain and external devices: the electrical activity of neurons is captured by electroencephalograms, which distinguish frequency spectra; then, neural activity is encoded and translated into commands that may operate devices such as high-tech prostheses, exoskeletons and typewriters (Schütz and Friedewald 2011 , 183)

11 The iPhone 6 s for instance features two cameras, a gyroscope, an accelerometer, a barometer, a proximity sensor, an ambient light sensor and a fi ngerprint identity sensor (Apple 2015 )

12 A report commissioned by the Council of Europe (CoE) has defi ned biometrics as “measurable, physiological or behavioral characteristics that can be used to determine or verify identity” (De Hert and Christianen 2013 ) Biometrics is therefore suited to human identifi cation through the use

technical capabilities of AmI while affective computing is a good example of both

“sensing and acting” in practice Combining a wide range of technologies that include sensors but also video surveillance and biometrics, affective computing is the use of ICTs to recognize and interpret human emotions and is a good illustration

of the relevance of sensors Processing information relating to emotions is a ble refi nement to the analysis of the physiological substrate of human behavior For instance, measuring physical variables (blood pressure or heart rate) should allow the inference of the underlying emotion (anger, fear or joy) (Gasson and Warwick

possi-2007 , 17) These technologies make it possible, among other uses, to label affective expressions – for example to determine if a smile is amused or polite (Hoque et al

2011 ) – or voice variations to identify emotions (Picard 2010 , 5) Using facial ognition to measure gender, age and mood and to offer products based on its analy-sis is already possible today: “[i]f you are an angry man of 30, and it is Friday evening, it may offer you a bottle of whiskey”, 13 says the head of marketing of a software vendor (Clifford and Hardy 2013 )

Embedded intelligence refers to the knowledge of machines underlying AmI Examples of embedded intelligence include the technical capability to build knowledge from observation of the user and act accordingly, “to infer the user’s context from whatever input is available and to reason about how to use data about the inferred context: in proactive suggestions to the user or in acting autonomously

on the user’s behalf” (Ahonen et al 2008 , 15) Aware of the diversity of accounts connected to “embedded intelligence”, 14 I will deal particularly with “machine

of body features such as fi ngerprint, hand, iris and face Traditional or fi rst-generation biometrics

is the paradigm to refer to mature technologies that have found large development in law ment or civil purposes, such as fi ngerprint and iris-based recognition More recently, second-gen- eration biometrics has been marked by two major trends: the processing of other types of traces and multimodality New processing of traces involves, for instance, analysis of motor skills such

enforce-as walking, voice and signature patterns, body signals such enforce-as electromagnetic signals produced by the heart and the brain, body odor and vein patterns, human-computer interaction such as key- stroke dynamics, facial recognition and soft biometrics, meaning the use of general traits such as gender, height and weight Second-generation biometrics is also multimodal as new systems take into account different biometrics simultaneously, differently from traditional biometrics where a single modality is deployed (Venier and Mordini 2011 , 116–121)

13 It is worth noting that more than 10 years ago in “Minority Report” – the SCI-FI fi lm adaptation

of Dick’s novel about a future year 2054 where a specialized police department arrests people based on foreknowledge – there is a scene where the main character is recognized through his iris

by a marketing computer system that offers him a beer – it seems we play harder in reality than in

fi ction

14 For instance, the idea of embedded intelligence is connected to artifi cial intelligence (AI) Intelligent agents emulate human intelligence: they are entities equipped with sensors and actua- tors and face challenges such as learning a model for prediction, creating a representation of the world, interacting in real time, competing and cooperating with other agents (Kleiner 2005 , 144)

AI can help AmI to accomplish tasks such as interpreting the state of the environment; representing information and knowledge associated with the environment; modeling, simulating and represent- ing entities in the environment; making decisions and planning actions; learning about the environ- ment, interacting with human beings and acting on the environment (Ramos, Augusto, and Shapiro

2008 , 16–17)

2.2 The Technology Within