Abstract

AI-Spotter is a Django-based web application built to address growing problem of AI-generated images being mistaken for authentic ones online. It allows users to upload suspicious images and evaluates them using an ensemble of five Python-based detection methods. These methods generate a unified confidence score estimating how likely an image is AI-generated. The system uses a Django 4.2.x backend with a Bootstrap front end, while the inference pipeline relies on PyTorch and Hugging Face Transformers, including a BLIP image-captioning model. Results are displayed with an overall confidence score and method-level indicators for clarity. The platform also includes authenticated REST endpoints for batch and real-time API access, secured with an API key. AI-Spotter follows ethical and security principles by emphasizing transparency, responsible AI use, file-type validation, and safe image parsing to reduce misuse and upload risks. Overall, it helps strengthen digital literacy by supporting more informed judgments about image authenticity online.