Читать книгу Data protection for the prevention of algorithmic discrimination - Alba Soriano Arnanz - Страница 5

The use of artificial intelligence tools and algorithmic solutions, both by private companies or actors and by public authorities, is nowadays a daily reality. There are, of course, some areas where these uses are much more common and developed. For example, the data collection and processing capacity of private companies in the digital economy, which include all the large intermediation platforms that have become global hegemons of the world economy, allows them to better understand their customers in general and, in particular, their needs and consumption patterns, in order to achieve greater efficiency, is already highly developed. This does not mean that computing capacity and the use of automated systems will not continue to increase in the future, for this increase or growth will certainly take place, thus increasing some already existing issues and making their analysis from a legal perspective even more important. We can also point to very important developments in the identification of patterns and strong correlations and inferences, which are very useful for some public activities, e.g. control and inspection duties. However, the use of such tools by public authorities, at least when taking administrative decisions with direct effects on citizens’ rights and interests, remains still relatively rare. Predictably, however, not for long.

There are countries that have advanced faster than others in this direction. The United States and the Anglo-Saxon world, where many of these developments have seen the light, often act as spearheads. Most of the problems that are already taking place in said countries regarding the legal framework applicable to the use of automated systems will soon be frequent in Europe. Many others are still to come in both parts of the world, but they will eventually arrive as the development of algorithmic tools becomes more advanced, almost on a daily basis.

Alba Soriano Arnanz, in this complete and informed work, carefully and analytically studies the full implications of these changes in a very exhaustive and complete manner. She does so, moreover, out of a very marked concern for the consequences that the use of these tools may end up having on our current policies designed to fight against inequalities of all kinds which, over the last half century, have become a hallmark in any truly advanced society. A concern that demonstrates not only her very pertinent sensitivity, but also a personal commitment so that this research work, this doctoral thesis I am honoured to present, is not only a theoretical exercise but also a tool of unquestionable practical utility. A practical tool which intends to help the creation of better artificial intelligence tools when they are used for the design and execution of public policies, but also at the achievement of a much better control of algorithmic tools when they are used by private companies.

This book, which is in part the result of Alba Soriano’s doctoral thesis, is therefore an excellent work from a theoretical point of view, but not only that. It contains a very refined study of how legal controls from the perspective of personal data protection have been developed in Europe and other legal cultures by a consolidated tradition of legal scholarship and how these tools are meant to be used in order to fight against cases of discrimination resulting from data processing technologies. For this purpose, this book contains an extraordinarily solvent, very clearly explained, exposition of the mechanisms and peculiar ways of functioning of algorithmic correlations and decisions mediated by artificial intelligences. But it also projects this study on the actual effects on reality that can be derived from those correlations and inferences made by AI artefacts.

For this reason, the main part of this work focuses on a detailed analysis of how data protection legal instruments that Western legal systems, and especially European ones, have to date been equipped can curb instances of discrimination and other risks to the fundamental rights of individuals and basic values and principles of democratic countries derived from the use of algorithmic tools. In this sense, the effort made by the author to clearly expose and explain in detail the complete possibilities offered by the European legal framework in the field of data protection to cope with those risks, based on the regulatory package embodied in the European General Data Protection Regulation of 2016, is highly praiseworthy and of great interest. Beyond that, Alba Soriano’s assessment does not fall short of exposing its inadequacies and shortcomings and describing specific practical problems that may arise from the limits, by their very nature, of the regulatory solutions that mainly derive from a personal data protection approach as the one that, at the moment, we have in Europe.

Alba Soriano’s commitment to offering a work with real practical utility also translates into a constructive criticism of this legal framework, which together with the exposition of its weaknesses is completed with a clever and meaningful outline of possible proposals for improvement, aimed not only at seeking a better adaptation of the current legal framework to the control of algorithmic discrimination, but also at trying to ensure better control of the use of algorithmic instruments in both the public and private spheres.

These characteristics of the work that I have the pleasure of presenting, fruit of impeccable legal analysis and theoretical depth but without ever losing sight of its vocation to be useful and to offer concrete solutions aimed to help to improve the effectiveness of public action in preserving and defending equality and the rights of the most disadvantaged social groups, define not only this book but also its author’s personality. Both combine legal rigour and expert knowledge with an exhibition of these values which are necessary for any jurist, and for any academic work, that aspires to be useful in a society such as the one in which we live today. To have been able to learn alongside Alba, and to follow the evolution of her work as academic supervisor up to her brilliant defence of these results as a doctoral thesis before a committee composed by Professors Juli Ponce Solé (University of Barcelona), Sofia Ranchordás (University of Groningen) and Gabriel Doménech Pascual (University of València) has been for me not only a privilege but a real pleasure. A privilege only surpassed by the opportunity that this whole process has given me to share my academic life with Alba, who is an even better person than a jurist and a brilliant scholar. Anyone who may have in their hands this book, the result of her work of the last three years, will be able to immediately perceive without great difficulty how hard it is to achieve this. But she does it.

Andrés Boix Palop

Administrative Law Associate Professor

University of Valencia - Estudi General