A Local, Continental (African) and International Overview of the Law as it relates (or tries to relate) to Artificial Intelligence (AI).

Abstract

In this article, the author performs a philosophical and a comparative analysis of the new phenomenon of Artificial Intelligence (AI) in a legal context. South Africa is in a a difficult position in adapting its legal system to such a radical change, having inherited both the Roman-Dutch law (a great part of its substantive law, especially Private Law), as well as English law (its entire Procedural Law as well as its Law of Evidence) from former colonial masters. One of the best international examples of specifically enacted AI law up to the present is that recently adopted by the European Union (EU). Unfortunately South Africa (RSA) will be forced to also have a close look at both the United Kingdom (UK) as well as the United States (USA) because the accusatorial system of procedure adopted by those countries is much closer to that applicable in the RSA. For the same reason, work done by the African Union (AU), being based mainly on the French inquisitorial model, is not ideally suited to former British colonies which are based on the British accusatorial model. Lay opinion on the future of AI, especially when allied with a powerful force (such as the Law), is also sharply divided. Perhaps because of science fiction hysteria, many citizens fear having their lives being over-regulated by inhuman machines. In the philosophical part of the present article some of these fears are addressed, mostly at the hand of  some commentary on the science fiction work of George Orwell, entitled "1984". Here Government, acting as "Big Brother", gains ingress to every household by means of a government-sponsored "two-way" . To the mind of the present author this is an excellent illustration of the conflicting claims for privacy and security that AI has brought about.