The Dark Side of Artificial Intelligence : challenges for the Legal System

• Author: Willem Gravett
• DOI: 10.25159/2522-6800/6979
• Published date: 01 October 2020
• Date: 01 October 2020
• Pages: 1-30

Article

Southern African Public Law

https://doi.org/10.25159/2522-6800/6979

https://upjournals.co.za/index.php/SAPL

ISSN 2522-6800 (Online), 2219-6412 (Print)

Volume 35 | Number 1 | 2020 | #6979 | 30 pages

© Unisa Press 2020

The Dark Side of Artificial Intelligence: Challenges

for the Legal System

Willem Gravett

https://orcid.org/0000-0001-7400-0036

Associate Professor, Faculty of Law, University of Pretoria

willem.gravett@up.ac.za

Abstract

The development of artificial intelligence has the potential to transform lives

and work practices, raise efficiency, savings and safety levels, and provide

enhanced levels of services. However, the current trend towards developing

smart and autonomous machines with the capacity to be trained and make

decisions independently holds not only economic advantages, but also a variety

of concerns regarding their direct and indirect effects on society as a whole. This

article examines some of these concerns, specifically in the areas of privacy and

autonomy, state surveillance, and bias and algorithmic transparency. It

concludes with an analysis of the challenges that the legal system faces in

regulating the burgeoning field of artificial intelligence.

Keywords: artificial intelligence; privacy; state surveillance; bias; algorithmic

transparency; algorithmic opacity; regulation

Gravett

2

Introduction

There can be no doubt that development of robotics and artificial intelligence (AI) has

the potential to transform lives and work practices, raise efficiency, savings and safety

levels, and provide enhanced levels of services in the short to medium term. Robotics

and AI promise to bring benefits of efficiency and savings not only in production and

commerce, but also in areas such as transportation, medical care, rescue, education and

farming. At the same time, they make it possible to avoid exposing humans to dangerous

conditions, such as those faced when cleaning up toxically polluted sites.

1

In the long term, however, the current trend towards developing smart and autonomous

machines with the capacity to be trained and make decisions independently not only

holds many economic advantages, but also raises a variety of concerns regarding their

direct and indirect effects on society as a whole.

2

It is the thesis of this article that AI is

creating a tangled web of legal issues that legal systems the world over will have to deal

with and resolve. The purpose of this article is not to offer comprehensive solutions, but

to raise awareness among legal scholars and practitioners of the most pressing legal

challenges presented by the increased application of AI. In particular, there are the

challenges to ensure privacy, the dissemination of factually accurate information, non-

discrimination, due process, transparency and accountability in decision-making

processes.

3

This article starts with a description of key terms; it then examines some of these

challenges highlighted above, specifically in the areas of privacy, the spread of

disinformation, state surveillance, and bias and algorithmic transparency. It concludes

with an analysis of the challenges that the legal system faces in regulating the

burgeoning field of AI.

Description of Key Terms

One of the main issues that must be faced when discussing the legal underpinnings of

technological innovations arises from the vocabulary used by those developing and

marketing these tools. Information technology (IT) professionals, like lawyers, have

developed a ‘somewhat dense and opaque lexicon’ that the uninitiated find too complex

to master.

4

It is important, therefore, to offer a general outline of the principal terms

1

European Parliament Committee on Legal Affairs, Report with Recommendations to the Commission

on Civil Law Rules on Robotics (PE582.443v03-00) (2017) 3.

2

ibid 4.

3

ibid 5.

4

Iria Giuffrida, Fredric Lederer and Nicolas Vermerys, ‘A Legal Perspective on the Trials and

Tribulations of AI: How Artificial Intelligence, the Internet of Things, Smart Contracts, and Other

Technologies Will Affect the Law’ (2018) 68 Case Western Reserve LR 751.