Journal of Simulation

Simulation and Generative Artificial Intelligence: From Model Construction to Machine Cognition

Generative Artificial Intelligence (GenAI) is rapidly reshaping research and industry practice. This includes Large Language Models (LLMs), Generative Adversarial Networks (GANs), and diffusion models. For the simulation community, GenAI could transform the entire modelling lifecycle. It can support building models from natural-language descriptions, speed up model coding, generate realistic synthetic inputs and scenarios, and give simulated agents context-aware cognitive capabilities. In turn, simulation offers controlled, stochastic environments in which generative systems can be tested, evaluated, and trained. This integration also raises significant challenges that demand rigorous scientific investigation. These include verification and validation, reproducibility, the fidelity of generated inputs and behaviours, and the ethics of AI-driven decision support.

Scope of this Special Issue

This special issue invites original contributions in theory, methodology, and applications, as well as systematic literature reviews and surveys, at the intersection of GenAI and simulation. It covers discrete-event simulation (DES), system dynamics (SD), agent-based simulation (ABS), and hybrid simulation (HS). We welcome work across three themes:

• GenAI for Simulation: generative models that support, automate, or augment the simulation modelling lifecycle.
• GenAI in Simulation: generative components embedded in simulation models, such as LLM-driven agents, learned behaviour models, synthetic populations, and generative surrogates or world models.
• Simulation for GenAI: simulation used to evaluate, test, train, or govern GenAI systems and the processes in which they operate.

Topics of interest include, but are not limited to:

• Input modelling and synthetic data: extracting inputs from unstructured sources; generative input models; fidelity of synthetic data and scenarios
• End-to-end modelling: natural-language-to-model workflows and GenAI assistance throughout the simulation lifecycle
• LLM-driven and generative agents in ABS and hybrid models
• Generative surrogates and metamodels for simulation optimization and uncertainty quantification
• Output analysis and decision support: interpreting and communicating simulation results to stakeholders and non-expert decision-makers
• Simulation-based testing, benchmarking, and training of GenAI systems
• Validation, reproducibility, and credibility of GenAI-augmented simulation
• Domain applications and lessons from practice: Real-world case studies from healthcare, manufacturing, logistics, defence, energy, commerce, public policy, architecture, planning, and environment
• Literature reviews and surveys that map the state of the art in GenAI and simulation

Submissions must demonstrate a clear and substantive contribution to simulation. Authors intending to submit a literature review are encouraged to contact the guest editors beforehand.