Xi Jinping Calls for Global AI Cooperation in Shanghai
Chinese President Xi Jinping delivered his first-ever keynote address at the opening ceremony of the 2026 World AI Conference (WAIC) in Shanghai on Friday, calling for international cooperation on AI governance and a commitment to open-source development. The address came as Chinese AI company Moonshot AI unveiled Kimi K3, the world’s largest open-source AI model with 2.8 trillion parameters, marking a major milestone in China’s artificial intelligence ambitions.
Speaking at the conference themed “Intelligent Partners, Co-creating the Future,” Xi emphasized that AI should become an international public good benefiting all humanity. According to Xinhua News, the president called on nations to seize the “rare historic opportunity” of AI-driven growth by encouraging open source, openness, collaboration, and sharing.
A Pivotal Moment in US-China Tech Rivalry
Xi’s address took place against a backdrop of escalating tensions between the world’s two largest economies. Just moments before the Chinese president spoke, US President Donald Trump delivered a prime-time national broadcast accusing China of orchestrating what he described as “the largest compromise of election data in history.”
The South China Morning Post described the dueling addresses as set to “further fuel escalating geopolitical and technology tensions between the two rival powers.” Xi’s speech emphasized cooperation and shared benefits, while Trump’s address framed China’s technological advances as part of a broader security threat — highlighting the starkly divergent approaches to AI governance emerging from Beijing and Washington.
Kimi K3: A New Benchmark in Open-Source AI
On the eve of the conference, Beijing-based Moonshot AI released Kimi K3, an open-source model with 2.8 trillion parameters built on a Mixture of Experts (MoE) architecture with 896 experts, activating only 16 per inference. The model features a 1-million-token context window and native visual understanding capabilities, optimized for software engineering, knowledge work, deep research, and multimodal understanding tasks.
According to Xinhua, a Moonshot AI spokesperson explained that “the larger the parameters, the higher the capability ceiling, providing smarter model performance. Nearly 3 trillion parameters mean the model can pack more knowledge and patterns into its ‘brain’ — understanding more, thinking deeper, and answering more accurately.”
The model’s performance in evaluations approaches that of frontier closed-source models, demonstrating China’s ability to compete at the cutting edge of AI research despite US chip export restrictions. Liu Tieyan, President of the Beijing Zhongguancun Academy, told Xinhua that “the vitality of technology lies in open flow. A batch of Chinese open-source large models is moving from ‘single-point亮相’ to ‘group breakthrough,’ providing new solutions and pathways for world AI development.”
World AI Cooperation Organization Established
A landmark development at the conference was the signing of the agreement to establish the World AI Cooperation Organization (WAICO), a new intergovernmental body with 29 founding member states. Chinese Foreign Minister Wang Yi signed the agreement on behalf of China, with UN Secretary-General António Guterres in attendance.
As reported by China News Service, the organization is an independent intergovernmental body under the UN Charter, headquartered in Shanghai. Founding members include Kazakhstan, Laos, Pakistan, Russia, and Indonesia, among others. The agreement states that WAICO aims to “promote international cooperation and global governance of AI, ensuring that AI develops in a beneficial, safe, and fair direction for the benefit of all humanity.”
China’s Vision for Global AI Governance
Xi’s keynote marks the culmination of China’s years-long push to position itself as a leader in AI governance. Since releasing its Global AI Governance Initiative in October 2023, China has steadily built a framework emphasizing AI for good, safety, and fairness. The country has implemented the AI Capacity Building Universal Benefits Plan, launched the “AI+” International Cooperation Initiative, and proposed the AI Global Governance Action Plan.
A key theme of Xi’s address was bridging the global AI divide. He called for helping Global South countries strengthen technical capacity, reflecting China’s positioning as a bridge for developing nations. According to research cited at the conference, Africa accounts for less than 1% of global data center capacity, and China has already implemented AI projects in Cambodia, Mozambique, Brazil, South Africa, and Djibouti.
Record-Breaking Scale
WAIC 2026 itself set new records, spanning three Shanghai venues with over 100,000 square meters of exhibition space. The event features more than 1,100 exhibitors, 3,000 exhibits, 300 global premieres, and over 140 forums. Attendees include nine Turing Award and Nobel laureates, alongside more than 1,400 Chinese and international guests.
What to Watch For
The simultaneous developments at WAIC 2026 signal China’s accelerating ambitions on multiple fronts. The combination of Xi’s personal endorsement of open-source AI, the record-breaking Kimi K3 release, and the establishment of WAICO represents a coordinated push to shape global AI architecture. Key questions remain: Will WAICO emerge as a genuine multilateral platform or a forum aligned with Chinese interests? Can Chinese open-source models continue to advance despite semiconductor restrictions? And how will the United States respond to China’s growing influence in AI governance?
As the conference continues through July 20, the answers to these questions will begin to take shape — with implications that extend far beyond the exhibition halls of Shanghai.