Using AI to combat AI-generated fraud: A semantic detector with blockchain integration

The rapid development of information technologies and their prevalence in the daily life of individuals, and especially the development of artificial intelligence in the last few years, have brought numerous benefits, facilitated and accelerated business activities, and enabled a wide range of modalities of use in different areas. In parallel with the development of artificial intelligence and its availability to a wide audience through increasingly numerous online tools, various modalities of its abuse have also been developed. Today, artificial intelligence has become the fuel and one of the main tools in various criminal activities on the Internet. Thus, various types of AI-assisted fraud have taken a dominant place on the ladder of high-tech crime today. One of the key questions is what measures we can take to fight against these types of abuse and whether this powerful technology can be used to fight against our own abuse. This paper presents the development and application of a semantic AI fraud detector as a conceptual model that uses NLP to recognize fraudulent messages and writes the results to the blockchain network to ensure data transparency and immutability, with suggestions for further integration and improvement.