Abstract:
Fake identities, including fully automated bots, humanoperated sockpuppets, and hybrid cyborgs, manipulate public discourse and amplify misinformation on social media. This review analyzes the technical evolution of fake identity detection from 2015 to 2026. As adversarial tactics advanced, defense mechanisms transitioned from feature-based machine learning to sequential deep learning and Graph Neural Networks (GNNs). Current state-of-the-art detectors face severe degrada-tion due to Generative AI, which produces synthetic personas capable of evading traditional NLP and structural filters.
Download full Lenght Paper......