ABSTRACT:The emergence of AI-generated deepfake videos — synthetic media fabricated using Generative Adversarial Networks (GANs), diffusion models, and related deep learning techniques — represents one of the most consequential technological developments in the contemporary information ecosystem. This paper investigates two interlocking dimensions of the deepfake phenomenon: the technical landscape of deepfake detection methodologies, and the social and psychological impact of deepfake proliferation on digital trust among social media users. Drawing on interdisciplinary literature spanning artificial intelligence, media studies, social psychology, digital sociology, and cybersecurity, this paper provides a comprehensive analysis of deepfake technologies and their societal impact. Key findings indicate that current detection methods remain in an adversarial arms race with generation capabilities. At the societal level, deepfake proliferation contributes to epistemic confusion, political polarisation, erosion of journalistic credibility, and a generalised crisis of digital trust. The paper concludes with a framework of technical, regulatory, platform-level, and media literacy interventions.

KEYWORDS: Deepfake Detection, AI-Generated Media, Digital Trust, Social Media Misinformation, GAN, Synthetic Video, Media Authenticity.

DETECTION AND SOCIAL IMPACT OF AI-GENERATED DEEPFAKE VIDEOS ON DIGITAL TRUST AMONG SOCIAL MEDIA USERS

PRANALI VISHNU MANE
SHREE RAM COLLEGE OF COMMERCE, BHANDUP