Detecting AI-Generated Artwork: An Ensemble Model for Enhancing Visual Authenticity in the Digital Age

Generative AI refers to a subtype of AI that involves creating diverse content such as text, images, and videos. Advances in AI targeted at enhancing content creation have allowed generative AI to grow successfully in producing such content. Generative AI technologies pose challenges in distinguishing between a real image and an AI-generated one due to potential misuse in fraud, manipulation, and harassment. This study proposes an innovative ensemble model architecture for the detection of artificial content in artworks. The proposed model architecture for detecting artificial content in artworks integrates deep learning models, including MobileNetV2., InceptionV3.,and ResNetV2, with custom layers appended to the base models. Transfer learning is leveraged by fine-tuning pre-trained ImageN et weights on the dataset tailored for this specific task. The ensemble model demonstrates high accuracy in classifying AI-generated art into real art or AI generated art, enhancing reliability in distinguishing synthetic content from human-created artworks. The ensemble model achieves an accurate classification of AI-generated art. This progress addresses the spread of misleading content and combats the distortion of facts within online platforms. This method, using neural networks, protects the originality of creative pieces in a time heavily influenced by machine-generated innovations.