The expansion of metaverse ecosystems and their convergence with artificial intelligence, virtual reality, and data-driven economies have fundamentally transformed patterns of victimization. In many instances, victimization is no longer the direct result of human conduct but rather emerges from the complex interaction among users, algorithms, and platform architectures. Within this context, the central research question is how the concept of victimization can be reconceptualized in an environment where both human and non-human actors simultaneously participate, and how legal responsibility can be attributed in such a setting. This study aims to explain the emerging nature of victimization in the metaverse, examine the role of algorithmic bias in generating harm, analyze existing legislative gaps, and evaluate the challenges associated with digital evidence. The research adopts a descriptive-analytical methodology based on documentary research and a comparative analysis of domestic and international literature in the field of digital criminology. The findings indicate that victimization within metaverse environments is structural, data-driven, and multilayered, with responsibility dispersed among platform providers, users, and algorithmic systems. The study further demonstrates that the inadequacy of traditional evidentiary frameworks in addressing fluid and immersive digital data, together with legislative deficiencies and the absence of safety-by-design standards, significantly contributes to the escalation of digital victimization. Consequently, an effective response to this phenomenon requires a transition from purely reactive approaches toward a preventive criminal policy grounded in security-by-design principles and intelligent regulatory oversight.