Abstract

Public sector investment portfolios are increasingly exposed to macroeconomic volatility, climate-related fiscal shocks, geopolitical instability, and structural revenue uncertainty. Traditional deterministic risk assessment models and static financial forecasting techniques are often insufficient for capturing nonlinear dependencies, tail risks, and dynamic policy interactions that influence long-term portfolio stability. This review paper synthesizes contemporary advances in artificial intelligence (AI)–driven financial risk simulation and conceptual modeling frameworks tailored to public sector investment management. It critically examines machine learning–based predictive analytics, deep learning–enabled scenario modeling, reinforcement learning optimization, agent-based simulation, and hybrid stochastic–AI architectures applied to sovereign funds, municipal capital improvement portfolios, infrastructure investment programs, and public pension systems. The review further evaluates how AI-driven Monte Carlo simulations, Bayesian networks, stress-testing algorithms, and ensemble forecasting models improve risk-adjusted capital allocation, debt sustainability forecasting, liquidity stress modeling, and intergenerational equity preservation. Conceptual models linking fiscal policy constraints, debt-service dynamics, asset lifecycle governance, and revenue diversification strategies are analyzed to assess systemic resilience under adverse shock conditions. Additionally, the study explores governance implications, model interpretability challenges, ethical considerations, and regulatory integration within public financial management frameworks. By consolidating empirical findings and theoretical perspectives, this paper proposes an integrated AI-enabled risk simulation architecture for enhancing transparency, stability, and long-term sustainability in public sector investment portfolios. The review contributes to the growing discourse on digital transformation in public finance by identifying methodological gaps, policy implications, and future research directions necessary for developing resilient, data-driven public investment ecosystems.