The global governance of artificial intelligence (AI) requires cooperation and coordination among international, national, and non-state actors. While existing scholarship has extensively explored the emerging AI regime complex, there is a notable gap in understanding public trust in the various stakeholders involved in global AI governance. This paper fills this gap through parallel surveys in the United States and China, two pivotal nations in AI development and regulation, locked in an ongoing geopolitical rivalry. I argue that in the evolving and contested domain of global AI governance, citizens will trust their own country more than others, with the trust deficit being larger for rival countries perceived as untrustworthy or hostile. Results show that respondents in both countries express the highest levels of trust in their own governments to lead global AI governance efforts, while trust in the other country is the lowest, below that of other stakeholders such as the European Union and tech firms. The findings also point to several factors that may mitigate the trust gap. In particular, those who hold more favorable views of AI’s societal impact and adopt a globalist perspective place higher trust in all stakeholders across the board, including the rival state. Overall, this study highlights the divergence in public trust toward global AI governance, which will need to be incorporated into efforts toward building a global governance framework, especially in the shadow of rising geopolitical tensions between major AI powers.