AI in Criminology Research: Mapping Methodological Shifts and Epistemic Risks

Over the past two decades, artificial intelligence (AI) has quietly but profoundly transformed the study of crime—from predictive algorithms used to identify policing hotspots to machine learning systems that analyse court records and sentencing patterns. While AI is no longer novel to many domain researchers, we still know surprisingly little about how these technologies are reshaping the discipline of criminology itself. This project will deliver the first comprehensive investigation into how AI—specifically methods in machine learning, deep learning, natural language processing, and large language models—is influencing what topics are studied, how research is conducted, and who is able to participate in this evolving field of knowledge production. We will analyse trends across twenty years of criminology research—examining published studies to track how AI adoption has shifted the topics researchers investigate and the methods they use. Are certain types of crime, like cybercrime or financial fraud, receiving more attention simply because they generate data that AI can easily process? Are traditional approaches like ethnography or case studies disappearing from top journals because they don't fit AI-driven research models? Our analysis will reveal these hidden patterns across all major subfields of criminology, from policing studies to prison research. Alongside this big-picture analysis, we will survey 100 leading criminology researchers globally—from early-career scholars to established professors—to understand how AI tools are affecting their day-to-day research. Do junior researchers feel pressured to adopt AI methods to be competitive? Are certain institutions pulling ahead because they can afford expensive AI systems while others fall behind? The survey will capture these human stories behind the technological transformation. A key focus will be understanding AI's impact on qualitative research—the in-depth interviews, community observations and archival work that have long been the backbone of criminology. Early evidence suggests these vital approaches may be declining as funding flows toward computational work. We'll investigate whether AI is complementing or crowding out these human-centred methods, and what this means for our understanding of complex social problems. The project will also explore whether AI is accelerating scientific progress in criminology—or potentially producing new knowledge in the discipline. While AI can analyse data faster than humans, it may also encourage researchers to chase trendy topics or rely on flawed datasets. We'll examine whether AI is helping criminology tackle important questions more quickly, or simply generating more studies of questionable real-world value. This work comes at a critical moment for the criminology discipline given the pressures of AI development. Police forces, courts and policymakers increasingly rely on criminology research to guide decisions affecting millions of lives. If AI is silently reshaping what we know about crime and justice, we urgently need to understand how—and ensure the transformation benefits society as a whole. By mapping AI's growing influence across criminology, this project will provide the evidence needed to guide this technological revolution responsibly. The findings will help: (1) Researchers make informed choices about adopting AI methods; (2) Funders and journals guard against AI-driven biases in what gets published and funded; (3) Policymakers understand how technological changes may affect the evidence base for crucial decisions.