GRouNdGAN is a gene regulatory network (GRN)-guided causal implicit generative model designed for simulating single-cell RNA-sequencing (scRNA-seq) data. It enables in-silico perturbation experiments, such as 'knocking out' specific transcription factors to observe gene reactions, and is used for benchmarking GRN inference methods. The tool bridges the gap between simulated and real biological data by generating realistic scRNA-seq datasets where genes are causally expressed under the control of their regulating transcription factors. It can synthesize cells under new conditions, allowing for in-silico TF knockout experiments. GRouNdGAN is valuable for understanding gene regulation, studying diseases like cancer where gene expression is disrupted, and aiding drug development by predicting cellular responses to genetic modifications.