The Federal Reserve Board published a research paper introducing CardSim, a payment card transaction simulation methodology intended to support fraud detection research where publicly available transaction data are limited. The paper describes CardSim as a flexible and scalable simulator designed to generate datasets that better reflect diverse and evolving fraud tactics without relying on sensitive real-world transactions. Key features include calibration to publicly available data, a Bayesian approach to linking transaction characteristics with fraud, and a modular structure implemented in a companion software package that can be updated as new evidence on payment trends and fraud patterns emerges. The outputs are positioned as inputs for testing and evaluating machine learning workflows, modeling approaches, and interpretability frameworks relevant to payment card fraud detection.