Abstract: A system and method for utilizing a generative adversarial model for training in a cybersecurity range is presented. The method includes: generating a plurality of scenarios based on a plurality of predetermined actions in a cloud computing environment; initiating a cybersecurity range, the cybersecurity range including a secure cloud computing environment and a plurality of workloads deployed therein; generating a prompt for a large language model (LLM) based on a template, a scenario of the plurality of scenarios, and an identifier of a workload of the plurality of workloads; and initiating an action in the cybersecurity range based on an output generated by the LLM in response to the prompt.