Detection of AI-Generated Sportsman Images using Deep Learning

Pavan Kumar M1, Rakesh j2
1 Department of Computer Science & Engineering K S Institute of Technology

2 Department of Computer Science & Engineering K S Institute of Technology

Abstract - The rapid advancement of Artificial Intelligence and deep learning technologies has significantly increased the creation of synthetic and manipulated digital media. Among these developments, AI-generated sports images and deepfake content have emerged as serious challenges for digital trust, sports journalism, athlete reputation, and online media authenticity. Fake sports images generated using Generative Adversarial Networks (GANs), diffusion models, and advanced image synthesis techniques can easily spread misinformation across social media platforms and digital news environments. Existing moderation systems mainly rely on manual verification processes, which are often slow, resource-intensive, and ineffective against large-scale synthetic media propagation.This paper proposes SGI-Verify, an AI-powered automated framework designed for detecting and removing AI-generated sports images from online platforms. The proposed system integrates metadata provenance verification, deep learning-based image analysis, automated moderation workflows, and compliance-aware takedown mechanisms within a unified architecture. The framework utilizes Convolutional Neural Networks (CNNs), forensic feature extraction, and C2PA metadata validation to identify manipulated sports images and verify digital authenticity. When suspicious content is detected, the system initiates automated moderation and takedown procedures while generating secure timestamped compliance logs for accountability and transparency.Key Words: AI-Generated Images, Deepfake Detection, Sports Image Verification, CNN, C2PA, Metadata Verification, Automated Moderation, Synthetic Media Detection, Image Forensics, Sports Media Security