Artificial intelligence in fine arts: A systematic review of empirical research

Abstract

Artificial intelligence (AI) tools are quickly transforming the traditional fields of fine arts and raise questions of AI challenging human creativity. AI tools can be used in creative processes and analysis of fine art, such as painting, music, and literature. They also have potential in enhancing artistic events, installations, and performances. In this systematic review, we investigated empirical studies on the use of AI in fine arts We gathered the data from three major bibliographic databases. After an initial search, we screened 723 articles based on pre-established inclusion criteria, resulting in 44 studies. Over half concerned visual arts, such as paintings and drawings, and one quarter concerned music. Experimental studies focused on human responses to AI art showed that people generally do not recognize the difference between human-made and AI-made art, but human-made art was valued more than AI-made art in some studies. The power of AI lies in the analysis of large-scale datasets. Case studies have reported development of an AI painter, DJ, performance artist, and music accompaniment for improvisations. AI tools have been applied to enhance consumer experience in online art shops and to provide virtual reality (VR) access to historical cathedrals. Fast development of AI will likely pose a challenge for the current concept and understanding of fine arts. AI challenging human creativity is one of the most powerful signs of the cultural and societal transformation stemming from AI.