Chinese President Xi Jinping used the BRICS summit in New Delhi to promote an open-source AI cooperation platform led by China and to encourage BRICS nations to join a Beijing-based World AI Cooperation Organization. This contrasts with China's cautious stance on restricting features of its domestic AI models amid concerns about security and control.
- Xi Jinping advocates China-led open-source AI in BRICS summit.
- World AI Cooperation Organization includes 29 countries, none from the EU.
- EU enforces AI Act standards with fines; China considers model usage limits.
What happened
At the BRICS summit held in New Delhi, Chinese President Xi Jinping proposed the creation of a community for open-source artificial intelligence, positioning China as the leader in fostering collaboration among BRICS nations to develop and apply large language models. He also formally invited the BRICS members to join the World AI Cooperation Organization, which was established by China in July and includes 29 countries, with headquarters in Shanghai. The organization notably excludes member states from the European Union.
Despite the invitation and Xi's promotion, the joint declaration from the summit did not explicitly endorse this initiative. Instead, it emphasized broader international cooperation on AI access that prioritizes safety, security, inclusiveness, and reliability. Meanwhile, within China, discussions continue over curbing overseas access to top-tier AI models such as Qwen, Doubao, and GLM, reflecting a cautious internal approach amid the external push for open collaboration.
Why it matters
This move highlights China's strategy to position itself as a key player in the global AI ecosystem by championing openness and cooperation among emerging economies, aiming to form a counterbalance to Western-led regulatory approaches. The World AI Cooperation Organization aims to build ties between member countries with shared governance over AI development, but it steering clear of including the EU signals geopolitical and regulatory divides in AI governance.
Meanwhile, the European Union has enacted its AI Act, which explicitly draws regulatory lines around openness and risk, exempting genuinely open-source AI models from certain compliance but still holding providers accountable for systemic risks. With the EU now empowered to impose fines up to 3% of global turnover for violations, the regulatory contrast with China—where AI governance remains less transparent and more fluid—could lead to divergent AI innovation and trade landscapes.
What to watch next
Observers will be closely monitoring whether the World AI Cooperation Organization gains influence and whether more countries, especially outside BRICS, join its ranks. How China balances this outward promotion of open-source collaboration with internal security concerns and potential restrictions on AI model capabilities will be crucial in shaping global AI supply chains and industrial cooperation.
At the same time, enforcement of the EU AI Act and its interpretation of systemic risk exclusions will set a regulatory precedent for other regions considering AI governance. The evolving policies in China and Europe offer a real-time comparison of governance models, with significant implications for AI developers, users, and international commerce.