Artificial Intelligence for Detecting Cyber Attacks in Deepfake & Identity Theft
Meghana Lokhande, Prajot Raut, Kiran Gawali, Mrudul Ahirrao, Abhishek Bhande · 2024
In today's era of digital world and evolving cyber threats, this research paper presents a unified exploration of innovative techniques harnessing the power of artificial intelligence and blockchain to combat deepfake attacks and identity theft. These intertwined challenges demand holistic solutions that transcend traditional boundaries. As the digital landscape is increasingly infiltrated by deepfake technology, concerns surrounding the authenticity of digital content are reaching a critical juncture. Deepfake attacks, capable of generating persuasive yet false imagery and videos, pose a grave societal threat. They undermine trust in media, perpetuate misinformation, and raise the specter of identity theft. Image processing techniques for deepfake detection aim to distinguish real from manipulated content by leveraging advances in AI. Meanwhile, the application of AI and machine learning in deepfake detection has yielded promising results, enhancing our capacity to discern authentic media from forgeries. The research converges on a proactive approach, introducing a pioneering framework that integrates AI and blockchain technology. This paper proposes an Artificial Intelligence-based protection framework, leveraging unsupervised pre-training techniques and Dense Neural Networks (DNN), to combat identity impersonation attacks, particularly the Clone ID attack directed at the Routing Protocol for Low Power and Lossy Networks (RPL). The research investigates the potential of blockchain, including Smart Contracts to combat the deepfake problem by verifying digital media's history and provenance.