Copyright Protection of Images Generated by Artificial Intelligence

Artificial intelligence has many uses in art fields such as music, poetry, and script creation, one of which is to use generative adversarial networks to generate new images or convert images. However, using generative adversarial networks to generate or convert images can also raise copyright issues. In the image input stage of artificial intelligence, including selection, programming, and training, it may involve works or software that others enjoy copyright. When outputting images, it is necessary to consider whether they constitute works and enjoy copyright. Generative adversarial networks involve diverse participants in the process of generating images, making it difficult to distinguish the creative contributions made by each party in the results. Therefore, the author's identity should be determined based on the analysis of the creative contributions from input to output.