Shutterstock

Superpower AI Rivalry Reshapes Global Technology Governance

07/13/2026

|

Simon Hutagalung | Eurasia Review

The Trump administration’s attempt to stop the import of Chinese artificial intelligence (AI) and Chinese restrictions on the export of frontier models are developments of great significance to the global digital landscape. Instead of an open and collaborative development of technologies for all of humanity, there is a new form of competition among countries to prevent others from accessing critical technologies to use against them. AI, once hailed as the future of humanity and a technological development to benefit everyone, is now viewed increasingly as a weapon of mass disruption that can be used to outcompete other countries.

The United States’ Executive Order also makes for a very interesting read. For the first time, the AI models developed by Chinese companies like DeepSeek, Qwen and GLM are characterised by the White House as nothing but systems for IP theft, propaganda and data mining. On the other hand, the United States also expanded its existing sanctions regarding the export of advanced semiconductors. This includes chips made by leading graphics card manufacturer Nvidia, such as the A100 and the newly released H100. These components are essential for the training of the so-called frontier models of AI. By banning their export, China’s ability to develop its own frontier models is severely curtailed. Chinese AI is therefore no longer welcome in the Western world. This also applies to all other countries, which are now also being implicated in strategic IP theft by working with Chinese AI.

China’s strategy mirrors the U.S. approach, with both countries operating in a spirit of strategic calculation. The restrictions placed by Beijing on the export of frontier models, including the Qwen2.5, DeepSeek V5, and GLM-6, all indicate that the country’s most advanced AI systems will be kept within its borders. Under a tiered system of regulation, the most basic models will be permitted for export, but those considered to be at the frontier of the field will be subject to a stringent security review before they are cleared for transfer abroad. As the United States attempts to starve China of semiconductors in a bid to strangle its AI ambitions, Beijing is working hard to achieve semiconductor independence to secure the future of its own AI systems.

The restrictions imposed by the two superpowers have, in the meantime, evolved into a restriction of imports and exports of frontier models. The global supply chain for AI of all kinds has become more restricted. The two major players in the world are cutting off the global supply of frontier AI models and, in this way, are in fact putting an end to cross-pollination of AI research between all nations. This will stifle innovation as it will lead to an overabundance of duplicate efforts, and the restriction of supply chains worldwide will furthermore have a negative effect on the efficiency of the global economy. Many countries are now forced into choosing a side between the Washington and the Beijing bloc and will become technologically dependent in the long run. Furthermore, they will no longer have a say in the worldwide standardisation of technology and in the corresponding and far-reaching rules that will be decreed by Washington and/or Beijing.

However, it is impossible to avoid this unhealthy competitive dynamic and therefore, a strategy of balanced engagement must be pursued. It is wrong for countries to be forced into alignment with either the United States or China and to be restricted from obtaining frontier AI models from around the world. While the US has its advantages in terms of being at the cutting edge of frontier AI models, China has its advantages in terms of open-weight models. A balanced engagement strategy, therefore, would be for countries to try to maintain open relations with both the US and China. The key is for countries to maintain open access to the AI models and systems of both superpowers, as well as those from around the world. In addition, countries can cultivate neutral partnerships with countries from around the world, such as those in the European Union, India and Japan. By fostering open, neutral partnerships, countries can create their own regional ecosystems and ensure that their own interests are protected, as well as create models and systems that are relevant to their own specific needs. The best way for countries to ensure that they are not becoming too dependent on either the US or China for their AI development and use is through open-source collaboration. This is the best way for countries to remain in control of their AI development and use, as it is a community-driven process and is therefore less likely to be controlled by either the US or China.

Artificial intelligence remains a shared global resource to foster human progress, should not be a means to an end of surveillance or a weapon of war in the hands of superpowers and countries globally. It will be critical to establish a robust framework of international ethics and security guidelines for the rapidly advancing technology. The ultimate protest or boycott of both countries’ models would be to reject them collectively and put both to task to rediscover and acknowledge the global nature of the resource they currently seek to use to further their individual interests in the world.

To read the full Op-ed as it was published by Eurasia Review, please click here.