Two giants – the United States and China – are locked in a competition over the future of AI. Their rivalry is disrupting technology markets and supply chains. It may divide the world into blocs that use different standards, political norms and philosophies of AI governance.

A third player, Europe, hopes to be more than a hapless bystander, and has some hard decisions to make about the balance that it wishes to strike between regulation and the promotion of innovation. Just to make matters more complicated, there is no consensus about the nature of this AI contest.

The most alarming view portrays the AI competition as an arms race, like the contest to acquire nuclear weapons. In this version, America and China are in a sprint to develop world-changing technologies, such as Artificial General Intelligence, which could bestow global dominance on the winner. Vice-President J D Vance has called for AI safety controls to be eased lest America loses this arms race and finds itself “enslaved to PRC-mediated AI”.

Another view is that AI is more like the internet. In this model, China’s priority is to develop a version of AI that is at once powerful and fully controllable by the Chinese Communist Party. Just as China has built its own internet behind a Great Firewall of controls, Chinese law declares that AI large-language models must not permit content that “damages the unity of the country and social harmony.” China is comfortable with algorithms that embrace state censorship and surveillance, and approaches to the profiling of criminal suspects that would be unconstitutional in many Western democracies. Chinese tech entrepreneurs see chances to sell such autocrat-friendly technologies around the world.

Finally, there are others who argue that China’s edge is its capacity to adapt new technologies to commercial uses and to develop practical applications, as if AI is more like electricity – a transformative force that will usher in new industries, public services and tools of governance. In this form of the contest, AI will be the fuel that powers a competition between two rival economic, industrial and political models.

If this contest is to end safely and productively, some sort of consensus will have to be reached between the major powers. An optimist would present the European Union as well placed to help shape that common understanding. A regulatory super-power with considerable leverage as the world’s last remaining open, super-sized consumer market, the EU stands mid-way between the individualism and risk-taking of the US and the smothering state control of China. It was a pioneer in AI governance with the AI Act, and its focus on precautionary, values-based regulation aims to build trust among anxious citizens and consumers. A sceptic would note that Europe lacks AI tech champions of its own and is under pressure to ease off on regulation in order to promote inward investment.