Written by Lykan van Gensen, Stellenbosch University
Leanne Manas is a well-known face on South African television. Towards the end of 2023, the face of the morning newscaster appeared elsewhere. They were fake news articles and fake ads featuring “her” to promote various products and get-rich-quick schemes.
It soon became clear that Manas had fallen victim to a “deepfake.” Deepfakes use artificial intelligence tools to manipulate images, video, and audio. Also, cutting-edge technical know-how is not required. Software like FaceSwap and ZaoApp are free to download, allowing anyone to create deepfakes.
Deepfakes were initially used in the entertainment industry. For example, a French actress was unable to film her soap opera role in person due to coronavirus restrictions, but she got the role thanks to deepfakes. In the medical industry, deep learning algorithms responsible for deepfakes are used to detect tumors through pattern matching in images.
However, such positive applications are rare. There is growing concern around the world about the impact of deepfakes on democratic elections. Recent reports suggest that deepfakes are on the rise in South Africa, and that South Africans are having a hard time identifying them.
It is therefore concerning that the South African government has yet to take legal steps to combat deepfakes, especially in light of national elections scheduled for later this year. I am a legal scholar specializing in sports law, with a particular focus on image rights. I am particularly interested in the recognition of individuals' portrait rights and the legal position when their portraits are misappropriated without their consent. This includes the use of deepfakes.
My LLD thesis considers the realities of digital media and the fact that many individuals, such as influencers, athletes, and celebrities, earn income by commodifying their images online. It argued that images needed clear legal protection. The promulgation of laws creates legal certainty in South Africa as it concerns the image of individuals.
international example
Various US states have already taken steps to combat deepfakes, primarily related to elections. For example, Texas became one of the first states to criminalize the use of deepfakes, especially when the deepfake content is related to political elections. It also recently passed a second bill targeting sexually explicit deepfakes. Therefore, deepfake videos are created for the purpose of harming political candidates or influencing election results, and sexually explicit deepfakes are created for the purpose of shaming political candidates without their consent. It is a crime to distribute or distribute.
Meanwhile, Maryland and Massachusetts have proposed bills that would specifically ban the use of deepfakes. Maryland plans to target deepfakes that can influence politics. Massachusetts already wants to criminalize the use of deepfakes for “criminal or tortious purposes.”
In 2020, California became the first US state to criminalize the use of deepfakes in political campaign advertising and advertising. AB 730 would make it a crime to publish audio, images, or video that creates a false and harmful impression about a politician's words or actions. Although the bill does not explicitly mention deepfakes, it is clear that AI-generated fakes are a major concern.
In 2023, the Governor of New York signed Senate Bill 1042A. This is aimed at banning the spread of deepfakes in general, not just those related to elections.
At least four federal deepfake bills are being considered. These include methods for identifying the output of generative adversarial networks and deepfake liability laws.
Protection of portrait rights
Currently, publicity rights are not recognized under South African case law or legislation. Portrait rights are legally different from copyright. The scope of protection provided by copyright alone is not sufficient to address the issue of deepfakes in court.
I advocate legal intervention to recognize individual portrait rights. Recognizing the rights to images protects them from unauthorized use. This includes not only the appropriation of personal images for commercial use, but also deep dives, whether related to elections, politicians, or malicious manipulation of personal images. This includes fighting fakes.
Portrait rights law is the key. can:
- clearly define your personal image
- Identify when image compromise occurred
- Provides image rights holders with legal remedies against unauthorized use.
All of these will help regulate the deepfake landscape. The malicious and deceptive nature of deepfakes can result in significant harm to image rights holders. It is time for the South African legislature to address these situations by providing the necessary protection to individuals.
Lykan Van Gensen, Junior Lecturer in Product Law; Stellenbosch University
This article is republished from The Conversation under a Creative Commons license. Read the original article.