DEEPFAKE AS A PHENOMENON OF THE MODERN INFORMATION SPACE

The problem of the circulation of unreliable (fake) information in mass communications is of critical importance for modern society. The development and accessibility of technology has negative consequences, including the possibility of using generative artificial intelligence (Gen AI) to create fake multimedia content (deepfake), disinformation and manipulation of public opinion. Modern technologies allow you to create and distribute false multimedia content, which is sometimes indistinguishable from reliable facts and is able to instantly spread on social networks (go viral), and then circulate in the information space for a long time. The key factors are the availability and speed of dissemination of news and entertainment information, a high degree of penetration of mobile devices and Internet connectivity, people’s trust in multimedia information and public figures. One of the triggers for the mass appearance of multimedia fake information after 2021-2022 in the field of technology is the commoditization (accessibility) of generative artificial intelligence (text2Image neural network Stable Diffusion, DALL-E). The available technologies of applied generative AI allow generating various types of multimedia content based on existing data - images, audio and video on any modern device (PC, smartphone), including connecting cloud computing power and multimodal LLMs available in the form of “chatbots” (ChatGPT, GigaChat, etc.). Generated and modified content is used in fraudulent schemes. The study describes the methodology for creating deepfakes and examples of their functioning in the modern information space, analyzes the risks of their use. Identifying fake content (deepfakes) created with the help of generative artificial intelligence is a difficult task that requires the development and implementation of effective algorithms and tools capable of detecting and distinguishing genuine materials from unreliable ones.The article analyzes the technological aspects of this process and current developments of foreign and Russian companies, such as Microsoft Video Authenticator, Intel FakeCatcher, “Vepr” and “Zephyr” software, technological solutions of Sberbank PJSC. The use of artificial intelligence technologies to create deepfakes poses a number of important ethical questions to society regarding truth, legality and moral acceptability, the right to privacy, copyright, as well as a potential threat to public safety. This requires the development and implementation of strict ethical standards and legislative regulations to prevent the abuse of these technologies, including mandatory watermark of Gen AI generated content. The article analyzes the positive experience of the Alliance in the field of Artificial Intelligence, which unites universities and technology companies in Russia. As a part of the organization’s activities, recommendations and ethical documents are being developed in order to regulate the use of AI technologies in order to prevent abuse and violations of ethical standards. The organization proposes to develop and implement approaches and technologies capable of detecting and distinguishing genuine materials from fake ones in order to help citizens and organizations critically evaluate information.