Biography

Isra Benchoufi is an emerging researcher in Artificial Intelligence and Cybersecurity, currently a Visiting Research Student at the Sensors, Clouds, and Services Laboratory (SCSLab) in the School of Computer Science at the University of Sydney. She is completing her MSc and Engineering studies at the National School of Artificial Intelligence (ENSIA), Algeria. Her research focuses on deepfake detection, reverse prompt engineering, and AI safety, with broader interests in deep learning, computer vision, and digital forensics.

Project Title:

Using reverse prompt engineering to discover deep fake images

Project Description

Deepfake technology poses a rapidly evolving threat by enabling the creation of highly realistic yet fabricated media that can convincingly manipulate images, audio, and video. Among its most concerning forms are AI-generated images produced by advanced text-to-image models such as DALL·E, Midjourney, and Stable Diffusion, which can generate photorealistic visuals rivaling genuine camera-captured content. This project aims to develop a robust solution capable of distinguishing AI-generated images from genuine camera-captured ones, while also determining whether synthetic images were created for benign or malicious purposes through reverse prompt engineering.