Opinion: US risks ceding AI rule-making leadership to China

Centre · 2Two opinion pieces published on 6 and 7 October 2026 argue that the United States is giving up influence over how artificial intelligence is governed internationally. 1 2 Centre · 1Newsweek's John Feng writes that China and the US should be fighting for the pen, but that America is barely in the running. 1 Centre · 1Feng says China created the World Artificial Intelligence Cooperation Organization (WAICO) this summer to draw in small and middle-income countries outside the US tech boom. He says many of its 37 members, including Cuba, Iran and Russia, are on America's AI cooperation blacklist. 1

Centre · 1The South China Morning Post's Denis Simon gives a different figure. He writes that the organisation, first proposed by Beijing in July 2025, became a reality in July 2026 when 29 countries signed the agreement establishing it in Shanghai. 2 Centre · 1Feng reports that Xi Jinping told Washington during a state visit that both nations had a responsibility to “manage AI for good”. He argues that for Beijing this means AI under the oversight of the Chinese Communist Party. 1 Centre · 1Feng says the US appears dismissive of WAICO. He says Pax Silica, the US-led coalition of 24 signatories set up in December 2025, focuses on supply chains and lacks a voice on AI governance or safety. 1

Centre · 1He adds that President Donald Trump believes whoever builds the best AI will make the rules, and that Trump has warned AI executives, including Anthropic's Dario Amodei, that regulation would let China win. 1 Centre · 1Feng cites Center for Strategic and International Studies analysts Chris Collins and Aalok Mehta, who say the US cannot credibly propose international standards for frontier AI without a coherent national set of regulations. They say a federal framework would give US diplomacy a model to offer other countries. 1 Centre · 1Simon's column starts from Trump's announcement of a voluntary agreement under which major AI companies will use internal controls, outside audits and board oversight. He notes that the White House calls the commitments “morally binding” but that the accord creates no comparable system of legal enforcement. 2

Centre · 1Simon contrasts this with the EU's AI Act, which imposes binding obligations on providers of general-purpose AI models. He says China combines domestic regulation with an active effort to shape the international discussion, citing its 2025 Global AI Governance Action Plan and its AI+ International Cooperation Initiative. 2 Centre · 1Simon predicts a patchwork of rules: American companies under a voluntary US framework, European users under detailed legal requirements, and other governments devising their own approaches. He says developing countries may find China's capacity-building message appealing. 2 Centre · 1On what to do, Simon does not call for copying Europe or China. He urges practical US-China channels on testing dangerous model capabilities, sharing information after serious incidents, and handling cybersecurity risks. He says this does not require a single global AI regulator, and he points to the UN's 40-member Independent International Scientific Panel on AI, appointed in February. 2

Centre · 1Feng also describes Chinese rules from the past six months on AI agents, human-like AI services and data for embodied AI, and quotes consultant Hugh Harsono on the risk that Chinese firms could gain influence over technical standards. Both pieces are opinion and reflect their authors' views. 1