Understanding the dynamics of criminal supply chains is crucial for addressing illegal activities while evading law enforcement. A central challenge is balancing the trade-off between security and efficiency, which profoundly impacts the structure of criminal organizations. However, it remains unclear whether differences in network structure correlate with reduced criminal efficiency. To address this, we present a stylized computational model rooted in game theory. The results illustrate how criminal organizations spontaneously emerge in environments with low perceived detection rates, where the benefits performing criminal activities outweigh the cost. This may inform law enforcement strategies on disrupting criminal organizations; The formation of criminal organizations can be attenuated by increasing the perceived risk of being detected, or decreasing the link density among members of the criminal organizations or disrupting the availability of required roles within the criminal organization.