Year: 2026
Type: conference-paper
Abstract: In recent years, the rapid development of artificial intelligence and deep learning technologies has significantly expanded the possibilities for creating and manipulating digital content. In this context, deepfake technology has emerged as a powerful tool capable of artificially modifying human faces, voices, and movements to generate highly realistic synthetic media. Although initially developed for entertainment and research purposes, deepfake technology has increasingly become associated with serious issues related to information security, political manipulation, fraud, and violations of personal privacy. This paper examines the origin, evolution, and advancement of deepfake technology based on modern artificial intelligence algorithms. The study analyzes the transition from traditional image manipulation techniques to advanced approaches utilizing Generative Adversarial Networks (GANs), autoencoders, convolutional neural networks (CNNs), and StyleGAN models. In addition, the paper discusses the widespread accessibility of deepfake creation tools, their social and political impact, and the growing risks associated with malicious use.
DOI: 10.65164/gzpqcm57
Author: Primbetov Abbaz1, Primbetov Aziz
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W7168282304
