How the AI race could play out for the US, China and the world

The global artificial intelligence (AI) race remains one of the most hotly contested spheres of international competition. Few would bet today on an eventual winner, but current trends point to three possible scenarios that we should consider.

In the first scenario, the AI world is dominated by the US and China. While the US still has the technological edge, China is closing in fast and vies with the US for leadership on many of the key dimensions of AI prowess.

According to Stanford University’s 2025 AI Index Report, a global benchmarking study of AI performance, the US leads in the number of major AI models and levels of investment, while China is ahead by a significant stretch in AI publications and patents granted. More worrying for the US, however, is that the technical performance measures for China’s AI models have dramatically improved in recent years and are within striking distance of US-based models.

What may prove even more decisive than technology in the longer term are the different economic models underpinning each country’s development of AI. China relies largely on a state-directed investment model that can mobilise resources, technologies and capital in pursuit of strategic national objectives. The prime constraint has been its access to high-end chips, but it is rapidly building capacity in this area and, moreover, the example of DeepSeek shows it is possible to creatively navigate such constraints.

In contrast, the US relies on a powerful nexus of technology titans, venture capital and government support. Yet worries are growing about potential fault lines in the US model, including the bubble-like appearance of the tech boom and the risks inherent in the use of complex debt mechanisms to fund data-centre expansion.

The second scenario, which might emerge amid a possible AI bubble collapse, is the rise of a tripolar AI world in which the European Union gains more sway for its ethics-driven approach to AI development.