Subscribe to industry newsletters

Search jobs

ChatGPT: Deepfake and copyright concerns

The creation and development of artificial intelligence (AI) has historically been accompanied with a warning label and understandably so. Stephen Hawking, amongst many, foreshadowed the concerns with AI and famously stated that development of AI "could spell the end of the human race".
Image source: © bpawesome –
Image source: © bpawesome –

The latest development in AI technology, which has raised widespread concern and criticism, is the creation of ChatGPT (generative pre-trained transformer), an AI computer program developed by startup company, OpenAI.

What exactly is ChatGPT?

ChatGPT is a language model AI computer program which generates output in the form of dialogue from text-based input received from the user. This output generated by the chatbot on ChatGPT, can take many forms including essays, research papers, even source code. Accordingly, a user can pose a question to the chatbot and receive a human-like dialogue in response thereto. The output that ChatGPT delivers is generated from a large dataset of text on the internet to which it has access. What contributes to the genius of ChatGPT is that the output is delivered in mere seconds. Remarkable right? Or too good to be true?

One of the main concerns with ChatGPT is its ability to generate deepfake text, which essentially means that deepfake creators can use this platform to create output which imitates human-like characteristics and work. Accordingly, the use of ChatGPT to generate academic articles, research papers or the like give rise to serious concerns regarding academic integrity and ethics.

Moreover, deepfake content, more specifically, in the form of video material can potentially have catastrophic consequences when consideration is given to political friction between countries such as Ukraine and Russia where a possibility might arise where a deepfake video surfaces purporting to be President Volodmyr Zelenskyy instructing his countrymen to surrender in the current war.

Authors vs owners

The more pertinent question to consider, from an intellectual property perspective, is who would be the author and owner of the copyrighted work created by such an artificial intelligence? South African copyright law distinguishes between the author and owner of a copyright work, as the author of the work may not necessarily be the owner thereof.

ChatGPT’s terms of use clearly assigns all its rights, title and interest in and to the output or content created, to the user. However, does ChatGPT even own the content which it so readily assigns? As a point of departure, the author is typically the first owner of any copyright work, as defined in the Copyright Act (Act 98 of 1978), but exceptions to this principal may apply as delineated in the Act.

The author of the output generated by ChatGPT will vary depending on the type of work created. For example, the author of a literary work is the person who first makes or creates the work, whereas the author of a computer program is the person who exercises control over the making of the computer program. Accordingly, the pertinent question is who would be the author of a literary work of such as an essay when the program generated the essay? The court, in Payen Components South Africa Ltd v Bovin Gaskets CC, drew a clear distinction between computer-generated work and computer-aided work. Computer-generated work is work created by the computer program, where the computer program was developed and designed by a human author. However, in the case of work generated by AI, the work is clearly not generated by a human author and control and direction is expended by AI.

Application of IP law

Clearly, the creation of ChatGPT is another innovative development in the world of AI but also creates further confusion relating to the appropriate application of IP law, especially copyright law. The identity of the author and owner of the output, despite an assignment provision contained in its terms of use, is still open to speculation and accordingly creates copyright works in which proprietorship cannot necessarily be identified.

This is indicative of the necessity in, amongst others, copyright law to adapt to the latest technology. More so in the sense of aiming at technological development rather than defining a specific circumstance.

About Stefaans Gerber and Alisha Muller

Stefaans Gerber is a Senior Associate - Head of Department (Intellectual Property) and Alisha Muller is a Junior Associate at Barnard Inc.

Let's do Biz