Detecting AI-Generated Images Using Convolutional Neural Networks (CNN) And Interpretation Through Explainable AI (XAI)

Abstract

ABSTRACT The rapid advancement of generative artificial intelligence has led to the widespread creation of highly realistic synthetic images. While these technologies have many beneficial applications, they also introduce challenges such as misinformation, identity fraud, copyright violations, and digital manipulation. This project proposes a Convolutional Neural Network (CNN)-based system to distinguish AI-generated images from authentic photographs. To improve transparency and trust in the model's decisions, Explainable Artificial Intelligence (XAI) techniques such as Grad-CAM, LIME, or SHAP are integrated to visualize the image regions that influence the classification. The proposed system aims to achieve high detection accuracy while providing interpretable explanations for each prediction.