Analysis of AI-Generated Art under Copyright

Smruti Sundar
University of Petroleum & Energy Studies, India

Volume III – Issue I, 2021

“The development of full artificial intelligence could spell the end of the human race. It would take off on its own, and re-design itself at an ever-increasing rate,” warned Stephan Hawking. What makes artificial intelligence superior to programs, is its ability to learn off of interactions and pre-existing data. This means that AI is capable of creating.

Copyright claims are often raised regarding the theft of art and understanding the difference between stolen work and inspired/derivative work is important. The incident of GANBreeder threw some light on the same where an AI created artworks not identical but extremely similar to that of other artists. It arose controversy as the original artists were enraged about the stolen artwork. Not all art created by AI had such striking similarities.

The intellectual property rights of the original creator can be at stake in cases of AI-generated artwork when one is unable to determine who the copyright belongs to. The AI does not have copyrights but if it has created artwork autonomously, how would its right, if any, be protected? And if the AI is given copyright, is there a violation of the rights of the developer or the user?

This paper will delve into the essentials to receive copyright and if AI-generated artwork would constitute under it. It will analyse the incident with GANBreeder where the AI-generated derivative/stolen works and explore the different approaches. This paper seeks to establish that AI can autonomously create artwork but cannot receive copyright despite fulfilling the essentials.

Keywords: Artificial Intelligence, Intellectual Property Rights, Copyright, Art by AI, GANBreeder.


 Download Full Paper