Please use this identifier to cite or link to this item:
https://www.um.edu.mt/library/oar/handle/123456789/70080
Title: | Regulating algorithmic profiling in the European Union : the interplay between EU data protection and anti-discrimination law |
Authors: | Rezki, Elysia (2020) |
Keywords: | Artificial intelligence Algorithms Data protection -- Law and legislation -- European Union countries Discrimination -- Law and legislation -- European Union countries |
Issue Date: | 2020 |
Citation: | Rezki, E. (2020). Regulating algorithmic profiling in the European Union: the interplay between EU data protection and anti-discrimination law (Master's dissertation). |
Abstract: | Algorithms increasingly determine whom access and opportunities are opened up for, and whom, by contrast, they close against. Utilising data mining tools, algorithmic profiling techniques are able to extrapolate highly sophisticated inferences concerning individuals or groups in the construction and application of profiles. The potential for algorithmic profiling to generate discriminatory outcomes has become increasingly contentious amongst policymakers, academics and wider society alike. This dissertation endeavours to contribute to the ongoing debate present in the European Union on how best to protect individuals from occurrences of discrimination generated by technology. Building upon existing work on algorithmic profiling, it examines specifically how the existing EU antidiscrimination and data protection legal frameworks may be applied in challenging discriminatory algorithmic profiling activities. The findings of this research relate to three pivotal areas: the applicability of the framework’s substantive prohibitions; the transparency tools enabling the individual to detect potential discrimination; and the role of public supervisory bodies and other legal entities in preventing algorithmic discrimination. The outcome suggests that the limitations of each field and remaining gaps of a combined approach mean the existing EU anti-discrimination and data protection frameworks are largely insufficient to adequately tackle algorithmic profiling. However, in acknowledging the unviability of an individual rights approach and futility of explanation rights, this thesis poses an enhanced regulatory focus on the pursuance of an ex ante ‘equality-by-design’ system, overseen by public supervisory bodies equipped with sufficient interdisciplinary expertise. Although the inherent contextuality of discrimination and equality mean even the most diligent of algorithmic models curated will still fall short in some manner, the pursuance of an ‘equality-by-design’ objective under careful consideration of the sought purpose and wider social context could ultimately prove the most effective path of regulation. |
Description: | LL.M.EUR.COMP. |
URI: | https://www.um.edu.mt/library/oar/handle/123456789/70080 |
Appears in Collections: | Dissertations - FacLawEC - 2020 Dissertations - MA - FacLaw - 2020 |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
20MLEC003.pdf Restricted Access | 1.12 MB | Adobe PDF | View/Open Request a copy |
Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.