Blockchain-Driven Defense Against Deepfakes on Social Media Platforms

Abstract

The rapid advancement of Artificial Intelligence (AI) has driven a surge in the spread of falsified digital content, particularly through deepfake technologies that generate highly realistic but fake videos, images, and audio. Such manipulated media can distort reality, damage reputations, and erode public trust. To address these risks, this paper proposes a blockchainbased framework designed to protect the authenticity and integrity of user-generated video content on social media platforms. The framework integrates two AI models: Multi-task Cascaded Convolutional Neural Network (MTCNN) for facial verification and Artificial Neural Network (ANN) for voice recognition, ensuring that both visual and audio identities are authenticated before publication. Once verified, blockchain records consent immutably using smart contracts, creating a secure log of authorization for accountability and transparency. This paper also explores blockchain’s role as a decentralized verification tool, examining deepfake detection methods, benchmarks, and the application of blockchain to enhance media reliability within a landscape increasingly susceptible to sophisticated manipulation.