Bridging Expert Insight and AI Reasoning: A Hybrid Systems Model of Iran's Fertility Dynamics

Population dynamics are inherently complex, shaped by nonlinear feedbacks among economic, cultural, health and governance systems. This study focuses on Iran's sustained subreplacement fertility and develops a hybrid modelling framework to construct a causal loop diagram (CLD) and integrates group model building (GMB), large language models (LLMs) and retrieval‐augmented generation (RAG) through a six‐step process: (1) initial dynamic hypothesis formulation; (2) expert‐driven CLD development; (3) AI‐driven CLD development; (4) model integration; (5) evidence anchoring using peer‐reviewed literature and (6) expert validation. The final CLD reveals how reinforcing mechanisms (e.g., education–modernity and economic confidence) interact with balancing constraints (e.g., childrearing costs, delayed marriage and institutional capacity) to sustain low fertility in Iran. The study demonstrates how structured human–AI collaboration can enhance transparency, theoretical grounding and policy relevance in systems modelling of demographic change, particularly in data‐limited and rapidly evolving contexts.