Exploring global contagion in artificial Intelligence, supply chain pressure, oil price, and economic policy uncertainty: insights from a bayesian TVP-SV-VAR model

Global crises ripple across technology, energy, policy, trade, and finance, yet the joint, time-varying links among Artificial Intelligence (AI), Supply Chain Pressure (SCP), Brent crude oil prices (COP), and Economic Policy Uncertainty (EPU) remain underexplored. This study provides the first integrated global analysis using monthly data from January 2014 to September 2024 and a Bayesian Time-Varying Parameter Vector Autoregression with Stochastic Volatility (TVP–SV–VAR) model to capture nonlinear and evolving spillovers. Robustness is demonstrated through unit-root and Johansen cointegration tests, Chow breakpoints around China’s AI strategy, the US–China trade war, COVID-19, and the Russia–Ukraine conflict, and by varying priors, lag lengths, identification, variable orderings, and sub-samples. Impulse responses are evaluated at one-, three-, and six-month horizons, representing short, medium, and long terms. AI lowers SCP and reduces EPU in the short horizon, with effects that fade at medium and long horizons, and its stabilizing role weakens during the Russia–Ukraine phase. COP generally stimulates AI, although this link turns negative during the Russia–Ukraine phase. COP’s impact on EPU is negative in the short horizon, mildly positive in the medium horizon, and negligible in the long horizon. EPU lifted COP before COVID-19 but reduced it afterward. SCP and EPU show heterogeneous, time-varying influences across variables. Policy priorities include accelerating AI adoption, diversifying and digitizing supply networks, and pursuing adaptive energy policies under credible governance. For emerging and transitional economies, resilience and restructuring hinge equally on digital capacity and energy security.