INFORMATION INFLUENCES ON INFOCOMMUNICATION NETWORKS WITH THE INVOLVEMENT OF ARTIFICIAL INTELLIGENCE

DOI: 10.31673/2412-4338.2025.038719

Abstract

The rapid development of information technologies has caused both positive changes in modern society and new threats to information security. Harmful information impacts on infocommunication networks are becoming increasingly sophisticated and dangerous, as attackers are constantly improving methods of influence, including by involving artificial intelligence (AI). Modern capabilities of AI technologies, in particular large language models (LLM), such as GPT-4 and BERT, and generative adversarial networks (GAN), allow solving complex problems of natural language processing (NLP). Cyberspace itself is changing, due to the use of AI technologies in both defensive and offensive operations. The relevance of the work lies in the need to identify areas of involving artificial intelligence in the formation of harmful information impacts on infocommunication networks and solving protection problems. The purpose of the study is to analyze modern information impacts on infocommunication networks using artificial intelligence to 

identify vulnerabilities, make informed decisions on ensuring information security, and improve protection methods. This study considers the possibility of using AI both for attacks on the infocommunication network and for protecting infocommunication networks. As a result of the analysis, it was found that not all information impacts involving AI are equally harmful. Among the potential threats and scenarios of information impacts involving AI that can negatively affect the information security of the infocommunication network, the most dangerous ones today are those that simplify phishing of users, generate human-like text, plausible voice messages, and video (deepfake) for more convincing deception of users. A promising direction is the active involvement of AI in the protection of infocommunication networks to counter cyberattacks, especially those carried out with the involvement of AI, ensuring the detection, analysis and response to cyberthreats in real time. Further research should focus on testing existing neural network models to assess the possibility of their application in solving problems of protecting infocommunication networks.

Keywords: information impacts; cybersecurity; LLM; artificial intelligence; models; cyber threats; cyber attacks; phishing; infocommunication networks