The ubiquitous nature of AI means that it operates across borders. However, countries around the world vary significantly in their capacities to develop and use AI systems independently and responsibly. Limitations due to inadequate data for use in AI systems, weak or outdated digital infrastructure, and low levels of AI literacy and capacity are likely to position many low- and middle-income countries (LMICs) at a significant disadvantage when it comes to leveraging the opportunities and protecting against the risks posed by AI. AI is also a resource-intensive process, requiring large investments and innovation, both at the state level and in industries. This often means it is difficult to tackle AI without support, and based on resources, it can create major gaps between countries.