Multi-agent Simulations and Phishing: Exploring Security in Cloud Environments

Abstract

This work addresses the impact of social engineering attacks, particularly phishing, which remains one of the main threats to cybersecurity in cloud computing systems. With the increasing adoption of cloud services by companies and individuals, the vulnerability to these attacks has grown, as social engineering directly exploits the human factor. The objective is to understand how these threats affect the integrity of cloud systems and data security, analyzing how attackers accumulate information and progress through different security levels. The methodology adopted for this purpose was the development of a multi-agent systems simulation model, where users, attackers, and defenders interact in a cloud environment with different security layers, considering variables such as user susceptibility and the attackers' abilities. The results obtained show that the success of attackers is directly related to user susceptibility and the gradual accumulation of information, highlighting the need for more effective and comprehensive training strategies.