ARTIFICIAL INTELLIGENCE-GENERATED ART AND THE QUESTION OF AUTHORSHIP

Abstract

Artificial Intelligence (AI) has quickly changed the artistic production by allowing machines to produce images, music, literature, and multimedia works that mimic the work of humans with regard to creativity. The latest developments of machine learning, especially deep neural networks, Generative Adversarial Networks (GANs), and diffusion-based models, have increased what computational systems can do: creating complex artistic patterns based on massive data. Such developments have also brought up critical theoretical, legal, and philosophical issues of authorship, originality, and creative ownership on AI-generated artworks. This paper looks at the technical underlying principals of AI generated art and discusses the processes by which algorithms discover stylistic tropes, generate visual shapes and respond to human intervention in user prompts and parameter adjustment. The paper also discusses the changing argument over authorship in AI-generated art, which takes into account programmers, dataset curators, artists, and end users advantages in the creative pipeline. Moral and cultural considerations are also outlined, such as the issues concerning intellectual property, cultural biasness in training data, and the possible repercussion to the conventional artistic careers. With the combination of the views of computational creativity, the digital humanities and the cultural policy, the study points to the transformative paradigm of human-intelligent systems collaborativity of creativity.