This study examined artistic expression by artificial intelligence(AI), which is one of the recently emerging topics in the human industries. By analyzing cases of artistic expression by AI up to the present, this study aims to ascertain the new possibilities and limitations presented by AI with regard to the artistic acceptance of science and technology. It also intends to provide the foundations for the future development of AI-based art. For this purpose, we first presented the concept of deep learning in addition to the direction and characteristics of how art is evolving in the age of AI. Based on this, we analyzed cases of AI-based artistic expression such as Deep Dream, The Next Rembrandt, and Aaron, making distinctions between the aspects of direct and indirect expression and based on the criteria of creativity and autonomy of action. This study’s case analysis revealed that machine learning systems built upon ANN, which mimics human neural activity, were able to imitate the information processing methods of the human brain-i.e., discovering patterns within the myriad data and identifying objects-such that they could learn diverse images from the 15th to the 20th century. We also found that they could use the vast accumulated data to freely draw pictures as they intended. Furthermore, we confirmed the need to study how the highly superior learning speed of AI, and the development thereof, would affect the arts in general. This study’s significance lies in that, by examining artistic expression through science and technology, it explores the changes in how the meaning of art-which is a relative attribute that changes with the times-is realized and expressed. Therefore, based on this research, we expect to expand the artistic expression of artificial intelligence that thought about the relation between art and artificial intelligence in various angles.