This research pioneers a novel approach to understanding how tax evasion behavior initially emerges in a population through advanced computational methods.

• Combines dual large language models with deep reinforcement learning to simulate realistic taxpayer decision-making
• Creates agent-based simulations that model the "big bang" moment when tax evasion begins to spread within populations
• Identifies key factors that influence tax evasion emergence, enabling better prevention strategies
• Demonstrates how AI can provide insights into complex socioeconomic security challenges

This research provides security professionals and policymakers with valuable tools to understand, predict, and potentially intervene in tax evasion dynamics before they become widespread societal problems.

Investigating Tax Evasion Emergence Using Dual Large Language Model and Deep Reinforcement Learning Powered Agent-based Simulation