Xi Urges Open-Source AI Sharing at Shanghai Summit as China Courts Developing Nations
Updated
Updated · The New York Times · Jul 17
Xi Urges Open-Source AI Sharing at Shanghai Summit as China Courts Developing Nations
3 articles · Updated · The New York Times · Jul 17
Summary
At a Shanghai AI conference, Xi Jinping cast open-source artificial intelligence as a "rare and historical opportunity" and said AI should be a "symphony of global collaboration," not dominated by one country.
Xi argued that wider sharing is needed to prevent "new historical injustices," framing China as a trusted partner for developing nations while advancing its bid to shape global AI rules.
China is still seen as trailing the United States in AI, but Chinese firms including DeepSeek, Moonshot and Zhipu are narrowing the gap with OpenAI's ChatGPT, Anthropic's Claude and Google's Gemini.
The pitch builds on Beijing's broader summit push, including a new AI cooperation body backed by 29 countries and offers of 5,000 training opportunities to partners in the Global South.
As Chinese AI models dominate on cost, can US firms adopt them without importing Beijing's rules and risks?
With its cheaper AI models, is China winning the race to write the world's AI rulebook?
As US and Chinese AI stacks diverge, are global businesses being forced to choose a technological side?
From 4% to 46%: How Chinese AI Models Overtook US Enterprise Workloads and Redefined Global AI Governance
Overview
From early 2025 to mid-2026, US enterprises rapidly shifted their AI adoption patterns, with Chinese-origin models rising from just 4.5% to consistently over 30% of enterprise token volume, peaking at 46.4% by mid-2026. This surge was driven by the significant cost advantages and competitive performance of Chinese models, leading to billions of tokens routed weekly. The trend marks a major transformation in the US enterprise AI landscape, as companies increasingly prioritize efficiency and affordability, reshaping global AI competition and raising new strategic and security considerations.