Social engineering with ChatGPT

Abstract

In this paper we are presenting a possibility of using ChatGPT for preparing environments for executing social engineering based attacks. Very soon after public launching, ChatGPT has shown excellent results in many different topics, answering some general or specific questions, generating code or preparing text templates of even texts on specific topics. Combining those possibilities and good preparedness of the system, it is possible to get everything that is needed for phishing or some other attack only in a few clicks and in a couple of minutes. This paper covers the scenario of creating phishing attack using ChatGPT, with an overview of social engineering attacks and their prevention in general.