Abstract
Drawing on rational choice and social learning theories, this study examines how social networks, subcultural immersion, and specific activities shape dark web actors' perceptions of risks and rewards. Using original 2024 survey data from 123 active dark web users recruited across English- and Russian-language forums and Telegram channels, we analyze how peer involvement, sustained participation in deviant subcultures, and interest in particular illicit activities influence perceived risks of detection and perceived benefits. Findings suggest that larger peer networks and deeper subcultural immersion may lower perceived risks and heighten perceived rewards. We also find important activity differences: interest in malware correlates with lower perceived risks, ransomware with higher perceived risks, and selling illicit goods with particularly high perceived rewards. These results underscore the fluid, socially constructed nature of risk and reward perceptions in clandestine spaces, offering rare empirical insight into the subjective mechanisms underlying cybercriminal choices and informing targeted prevention strategies.