Abstract: A method and system for defending against finance-related phishing attacks is presented. The method includes detecting when a message is a finance-related message based on contents of the message, wherein a finance-related message is a message that references at least identifiable financial elements; for each finance-related message, identifying a type of finance-related malicious activity, wherein a type of finance-related malicious activity is based on characteristics of the finance-related message; performing, using at least one AI agent from a network of AI agents, in-depth analysis of the contents of the message, wherein the at least one AI agent is specialized to detect the finance-related malicious activity of the identified type; and when a finance-related malicious activity is detected in the message as a result of performing the in-depth analysis, causing execution of a mitigation action.