Leveraging Convolutional Neural Networks for AI Artwork Classification

The emergence of AI-made art is pioneering in the creative landscape, yet it challenges quality and authenticity across many image-generation platforms. This paper aims to develop a new AI art detector capable of checking the authenticity of artwork. The developed detector will be designed in such a way that it can evaluate essential factors regarding originality, coherence, and aesthetic appeal. This is achieved by using deep learning techniques through training over a wide variety of datasets, from those generated by AIs and humans within the context of artwork. This process will automatically identify markers that are unique to content generated by AIs. With this, the findings will provide insights into the strengths and limitations of existing AI art models, leading to improvements in the future. Commercial art production, content moderation, and intellectual property management are just some of the possible fields in which our tool can be applied by maintaining quality levels for AI-generated artworks.