Confronting with the algorithm bias risks, if blockchain will provide new chance or challenge from data protection law perspective?

Li, Z. (2021) Confronting with the algorithm bias risks, if blockchain will provide new chance or challenge from data protection law perspective? Dublin Law and Politics Review, 2, pp. 43-67.

Full text not currently available from Enlighten.

Publisher's URL: https://heinonline.org/HOL/LandingPage?handle=hein.journals/dublpr2&div=19&id=&page=

Abstract

With the increasing concerns about the risks of automatic decision-making algorithms, the GDPR is expected to play a greater role in monitoring data controllers and processors. However, as algorithm complexity has increased, biases become more complicated, easier to hide and harder to control or measure. Meanwhile, the emergence of new technologies such as distributed ledger technology (DLT) bring new opportunities, but also challenges the legal basis of the GDPR and affects the implementation of EU data protection law. The most common type of algorithmic bias may be online behaviour targeting. From an economic and social efficiency perspective, there are benefits to this marketing strategy; however, the consumers' rights as data subjects are infringed. As decentralised technology progresses, some suggest that DLT could become a means to achieve the data protection law's objectives. Because data in blockchain systems are encrypted, and consensus algorithms establish a decentralised system to ensure transparency and data integrity. However, there is also a conflict between DLT and the GDPR. First, the GDPR is based on the current digital economy model that there be at least one natural or legal person as the data controller to undertake legal obligation. However, DLT is a shared, synchronised and decentralised digital database in which data are replicated and stored on multiple nodes. The second tension is that blockchains are append-only ledgers which means that data integrity can be ensured, but it is difficult to modify and delete existing data. These characteristics of the blockchain is incompatible with right to be forgotten in GDPR. This paper firstly defined algorithmic bias and its practical implications by using two examples. Then a brief overview of DLT will be provided. This article, furthermore, analysed what blockchain can bring to data protection law against algorithm bias. It is concluded that blockchain can render users to really control over their personal data, although certain drawbacks and incompatibilities with GDPR exist. Finally, the article will analyse how the conflict between blockchain and the GDPR can be reconciled.

Item Type:Articles
Status:Published
Refereed:Yes
Glasgow Author(s) Enlighten ID:Li, Mr Zihao
Authors: Li, Z.
College/School:College of Social Sciences > School of Law
Journal Name:Dublin Law and Politics Review
Publisher:Dublin City University

University Staff: Request a correction | Enlighten Editors: Update this record