US, China cooperate on global AI governance despite intensifying competition

The intensifying competition between the US and China in artificial intelligence (AI) continues alongside efforts to cooperate on governance rules, monitor AI-powered cyberattacks, and prevent frontier models from being controlled by non-state actors.

Two distinct approaches to global AI governance are emerging from the US and China, while a third approach is coming from the UN, according to a recent report by the Center for Strategic and International Studies (CSIS).

The number of partner countries in the US-led Silica Initiative, focusing on AI and semiconductor supply chains, reached 24 in June, while, on the other side, some 29 countries joined the rival Chinese-led World Artificial Intelligence Cooperation Organization (WAICO) as founding members, the report published on Sept. 21 showed.

Pax Silica was made up of participants primarily from high-income economies and the US’ security partners in Europe, the Indo-Pacific, the Middle East, and Latin America. WAICO members are, in contrast, from low- and middle-income countries, with Kazakhstan being the only country participating in both initiatives.

The key AI agenda items between the US and China mostly focus on cooperation in tracking AI-powered cyberattacks, China’s access to advanced American models, Washington’s restrictions on open-source Chinese models, and preventing state-of-the-art frontier AI models from falling into the hands of non-state actors.

During Chinese President Xi Jinping’s visit to Washington last week, talks between Xi and his American counterpart, Donald Trump, focused on developing AI in a human-centered and controllable way, as well as potential risks.

The CSIS report stated that the dialogue between the two parties could focus on developing common definitions to serve as a foundation for future discussion, such as determining which systems can be classified as frontier models.

One of the key elements of technological competition between the two countries is access to high-performance chips used in developing models.

US export controls on advanced technologies make it difficult for Chinese AI labs to access such chips and other computing resources, while China is moving toward developing its own semiconductor and AI capabilities.

The report highlighted that there is still no established mechanism for the parties to verify model security compliance, which stands in the way of potential AI security regulations.

US-China issues ‘no longer limited to trade deficits or tariffs’

Arzu Al, a political economy professor at Marmara University, told Anadolu that the meeting between Xi and Trump in Washington should be regarded as a strategic balancing act to manage the costs of competition, noting that the core issue between the two nations is no longer limited to trade deficits or tariffs.

“AI, advanced semiconductors, data infrastructure, energy capacity, and critical minerals became integral parts of economic power, national security, and foreign policy capabilities,” she said.

“Xi emphasized that AI must remain under human control and that its risks and benefits should be addressed together, while Trump supported the continued dialogue and cooperation in the area – this by no means constitutes the end of their technological competition, but in fact, it shows that the competition is intensifying in some areas while other areas remain open to controlled cooperation,” she added.

Al stated that the structure we’re seeing today should not be explained through the lens of classic Cold War bloc formation between the countries, but through a “controlled techno bloc formation.”

While the US and China have yet to fully sever their economic ties, the two countries also do not want strategic interdependencies to limit their potential.

“What is decisive in the technopolitical order is not so much the interdependence but rather the extent to which the interdependence is asymmetrical and whether it can be weaponized to pressure one side,” she said.

Al noted that Taiwan coming up during the meeting showed that structural disagreements between the US and China on security matters persist.

“Maintaining serious competition in security and technology while cooperating on AI at the same time is not a contradiction – it is precisely the nature of the current relationship between the two countries,” she said.

“This is also precisely where the real significance of this meeting lies, as the competition between Washington and Beijing is not mutually exclusive, and both competition and cooperation will exist simultaneously in this new era – the struggle to grow technological capacity and access critical inputs will continue to be one of the fundamental axes of US-China relations,” she added.