IGF 2023 WS #458 Where is the Author? AI-generated images, art & authorship


Artificial Intelligence (AI) & Emerging Technologies
Chat GPT, Generative AI, and Machine Learning
Future & Sustainable Work in the World of Generative AI

Organizer 1: Terezinha Alves Brito, NEDDIG - UFRN and GEDI - UFPR
Organizer 2: Julio César Márquez Ariza, Independent
Organizer 3: Pedro de Perdigão Lana, ISOC Brazil
Organizer 4: Umut Pajaro Velasquez, Internet Society

Speaker 1: Julio César Márquez Ariza, Civil Society, Latin American and Caribbean Group (GRULAC)
Speaker 2: Pedro de Perdigão Lana, Technical Community, Latin American and Caribbean Group (GRULAC)
Speaker 3: Raashi Saxena, Civil Society, Asia-Pacific Group
Speaker 4: Stella Anne Ming Hui Teoh, Civil Society, Asia-Pacific Group


Umut Pajaro Velasquez, Civil Society, Latin American and Caribbean Group (GRULAC)

Online Moderator

Terezinha Alves Brito, Civil Society, Latin American and Caribbean Group (GRULAC)


Theorose Elikplim Dzineku, Civil Society, African Group


Round Table - 60 Min

Policy Question(s)

A. How is it possible and reasonable to apply the concept of authorship to the images generated by an AI?
B. What is the space of the cultural baggage and creation process in the construction of AI as an author?
C. How can be involved privacy, ethics, accountability, fairness, and trustworthiness in the analysis of AI-generated images as a work or a product?

What will participants gain from attending this session? Participants will gain a broader understanding of the authorship concept, not just legally speaking (for copyright reasons) but also philosophically, and why this discussion is even more relevant than ever in the era of generative AI. We hope that participants will have new insights about the relationship between the author and their work and how it shapes our modern comprehension of art - artistic works are not economically and socially valued just by their inherent characteristics and aesthetics, but also because of the meaning they carry and their cultural worth, something that is closely connected to the motivations and goals of the person or people who created them.


Authorship is going through a crisis compared to what was traditionally understood. This discussion is central thanks to applications of Artificial Intelligence (AI) - from the creation of avatars to the development of artwork based on datasets that associate text to images, used for, among others, education, art, marketing, design and research.
These uses raise new ways of mediation between people and technology, beyond what happened with the emergence of the informational society. The question arises about the author of the resulting image and, in the same sense, what changes this type of use would generate to the concept of author and who would be inflected by this. With this in mind, we mention three concerns:
- Artists: the loss of the creative and artistic process, competition risks, as well as copyright and property rights.
- Privacy advocates: the use of personal data to train AIs, as it may be problematic considering that a person's data has multiple purposes.
- Ethicists: The application does not know right from wrong, and can easily be tricked into producing any utilization of any image shown to it, or can substantially harm an entire category of the creative economy.
Assuming there is a private company with a product, a programmer for the AI, a data set that makes it work, and users who generate new images, then we ask where is the author, not only from the legal perspective. What are the economic and social aspects involved in the development of these applications? We propose to confront social and technological concepts to investigate the transformations of the author and approach the potential new scenarios of mediation between the medium, the message, and the users, in order to understand the cultural production of these AI in economical, philosophical, and legal terms.

Expected Outcomes

This session seeks to mobilize awareness and promote the exchange of ideas around the issues surrounding the protection and commercialization of works generated by artificial intelligence, as well as the IA application themselves. As this is becoming an agenda of new regulations around the world, we consider it important that people participating in these discussions have a broader understanding of generative AIs that do not take as a starting point works isolated from their cultural contexts. As a concrete output of the session, we aim to produce a document listing a set of recommendations that policymakers should take into consideration when regulating generative AI focused on artistic works.

Hybrid Format: The structure of the session is as follows: (i) Presentation of the topic ; (ii) Examples of images generated by the different AI algorithms; (iii) Questions in conceptual opposition to open the discussion; (iv) Conclusions from the discussion.
We will promote hybrid participation by (i) alternating questions between online and on-site audiences; (ii) actively inviting some people interested in the theme to participate online; (iii) giving an active role to the online moderator who will be in charge of conducting the presentation; (iv) Mentimeter will also be used to allow for anonymity in questions and commentaries.