This study employs the MS‐GARCH‐EVT‐vine copula model to examine changes in the dependence structure and risk spillovers among global stock markets during the COVID‐19 pandemic. Our results indicate that the dependence structure of global stock markets exhibits intercontinental clustering characteristics. Specifically, the Hong Kong, French and US stock markets serve as the central nodes in the Asia‐Pacific, European and American regions, respectively. Furthermore, the COVID‐19 pandemic has reduced the number of stock markets directly linked to central nodes and exacerbated the synchronized decline in global stock markets. Additionally, the COVID‐19 pandemic has increased risk spillovers among global stock markets outside China, altering the direction of intercontinental risk contagion. These findings are significant for policy makers to prevent cross‐border risk spillovers and for investors to enhance their risk management strategies.