China challenges US AI dominance with open-source technology and new global body

Chinese President Xi Jinping used the opening of the World Artificial Intelligence Conference in Shanghai on July 17, 2026, to present China as a leader of a more open international artificial intelligence system, challenging technology restrictions and governance frameworks associated with the United States.

The Chinese leader called for artificial intelligence development to become a global cooperative effort rather than a field dominated by one country. He promoted open-source technology, greater access for developing economies and international mechanisms intended to keep advanced systems under human control.

China also announced practical commitments aimed at expanding its influence across the Global South. Beijing will provide 5,000 artificial intelligence training opportunities to developing countries over five years and offer 30 countries access to a Chinese-developed meteorological system that uses artificial intelligence for early warnings.

The speech followed an agreement signed by 29 countries on July 16 to establish the World Artificial Intelligence Cooperation Organization. The proposed intergovernmental body will be headquartered in Shanghai and is intended to increase the participation of developing economies in artificial intelligence governance.

The diplomatic message was reinforced by China’s domestic technology advances. Huawei Technologies showcased its Atlas 950 SuperPoD computing system at the conference, while Chinese developers promoted open models as lower-cost alternatives to largely proprietary systems developed by leading United States companies.

Why did Xi Jinping use the Shanghai artificial intelligence conference to challenge United States influence?

Artificial intelligence has become a central area of strategic competition between China and the United States because it combines economic power, military relevance, industrial productivity and control over emerging international standards.

Washington has restricted Chinese access to advanced semiconductors, chipmaking equipment and selected technologies that could support artificial intelligence development. The United States maintains that those measures are required to protect national security and prevent advanced computing systems from strengthening Chinese military capabilities.

Beijing argues that the restrictions extend national security policy beyond reasonable limits and prevent other countries from sharing in technological progress. The July 17 address placed that objection inside a broader diplomatic argument about inequality.

Rather than presenting the dispute only as a contest over chips, China framed artificial intelligence access as a development issue. The position is designed to resonate with countries that lack computing infrastructure, training data, specialised talent and the financial resources needed to build advanced models independently.

The strategic value of this message is considerable. Many developing countries do not want to choose permanently between Chinese and American technology systems. They want affordable tools, training, infrastructure and the ability to participate in decisions that will affect employment, education, healthcare and government services.

By offering cooperation rather than simply criticising United States restrictions, Beijing is attempting to convert its technological position into diplomatic influence. Training programmes, meteorological tools and application centres can create long-term institutional relationships with governments that might later support Chinese positions in international negotiations.

The Shanghai conference therefore marked a shift from defensive language about export controls towards a more ambitious effort to shape the global artificial intelligence order.

What would the World Artificial Intelligence Cooperation Organization add to global governance?

The agreement to create the World Artificial Intelligence Cooperation Organization gives China a permanent institutional platform for its governance agenda.

Twenty-nine countries signed the founding agreement in Shanghai. Participants included Russia, Pakistan and Kazakhstan, while the United States was not part of the initiative.

The proposed organisation is intended to operate as an intergovernmental body headquartered in Shanghai. Its stated purpose is to support international cooperation, reduce technological inequality and give developing countries a greater role in writing rules for artificial intelligence.

The institution could become important because existing artificial intelligence governance remains fragmented. The United Nations, Group of Seven, Organisation for Economic Co-operation and Development, European Union and several regional organisations have developed principles or regulations, but no single system governs the technology globally.

China is positioning the new organisation as a development-focused alternative. Its approach emphasises state sovereignty, infrastructure access, technology sharing and the economic benefits of artificial intelligence.

Western governance models generally place greater emphasis on privacy, human rights, accountability and safeguards for high-risk systems. The two approaches are not mutually exclusive, but they differ in political priorities and regulatory philosophy.

The emerging competition is therefore about more than technical standards. It concerns who defines acceptable uses, which governments participate in rulemaking and whether artificial intelligence oversight should be led by democratic values, national sovereignty or universal institutions.

The new body will need to prove that it can produce more than declarations. Its significance will depend on funding, membership growth, technical programmes and whether participating countries treat its standards as credible.

A governance organisation dominated by China could face criticism that it lacks independence. A body capable of delivering affordable infrastructure and training, however, may still attract countries frustrated by limited access to Western systems.

How do China’s training and meteorological commitments support its Global South strategy?

China’s promise to provide 5,000 artificial intelligence training opportunities gives the Shanghai agenda a practical dimension.

Advanced technology ecosystems require more than software. Governments need engineers, data specialists, regulators, cybersecurity professionals and public officials capable of evaluating systems before they are deployed.

Training programmes can build those capabilities while establishing professional networks linked to Chinese universities, companies and public institutions.

The meteorological initiative has an equally clear development purpose. Thirty countries will be offered access to a Chinese artificial intelligence system designed to improve weather forecasting and early warnings.

Many developing economies are highly exposed to floods, droughts, tropical storms and extreme heat but lack dense observation networks or advanced forecasting capacity. Better warning systems can support evacuations, agriculture, water management and disaster planning.

China can therefore present artificial intelligence cooperation as a response to immediate public needs rather than an abstract technological competition.

Beijing also plans to expand partnerships with the Association of Southeast Asian Nations, African Union, League of Arab States, Community of Latin American and Caribbean States, Shanghai Cooperation Organization and BRICS countries.

These institutions cover regions where China has already built significant influence through infrastructure finance, trade and the Belt and Road Initiative. Artificial intelligence could become another layer in those relationships.

The strategy may also create dependence. Countries using Chinese systems could rely on Chinese cloud services, hardware, maintenance and technical standards.

Governments accepting the offers will need to assess data protection, cybersecurity, procurement transparency and the long-term cost of operating the systems after initial support ends.

The programmes nevertheless strengthen China’s argument that artificial intelligence governance must include countries that have historically consumed technology without participating equally in its design.

Why are open-source artificial intelligence models central to China’s diplomatic pitch?

Open-source and open-weight models allow developers to inspect, adapt or deploy parts of an artificial intelligence system without relying entirely on a closed commercial service.

Chinese developers have used this approach to increase adoption and challenge the assumption that the most capable systems must come from United States companies.

DeepSeek demonstrated that Chinese models could gain international attention through strong performance and comparatively low operating costs. Moonshot AI used the Shanghai conference to unveil Kimi K3, which the company described as the world’s largest open model by parameter count.

Open models are attractive to governments and companies that want greater control over data, deployment and customisation. They can be adapted for local languages, public services and industries that may not be profitable priorities for major American providers.

The model also fits China’s diplomatic narrative. Beijing can present open technology as a public good while describing proprietary systems and export restrictions as barriers that preserve technological inequality.

Openness does not eliminate risk. A model may be available for adaptation while still requiring expensive computing infrastructure. Open systems can also be modified for cyberattacks, surveillance, disinformation or other harmful purposes.

There are further questions about the meaning of open-source in artificial intelligence. Developers may release model weights while withholding training data, methods or safety documentation.

China’s own technology environment is also heavily regulated. Domestic providers must comply with political, security and content rules imposed by Chinese authorities.

The global debate will therefore examine whether Beijing’s open-source message represents genuine technological access, a strategic export model or a combination of both.

How does Huawei’s Atlas 950 SuperPoD reflect China’s response to semiconductor restrictions?

The Atlas 950 SuperPoD represents China’s effort to build large-scale artificial intelligence computing systems using domestic processors and networking technology.

Advanced artificial intelligence models require enormous computing power for training and operation. United States restrictions have limited China’s access to the most advanced chips supplied by Nvidia Corporation and to equipment used for manufacturing leading-edge semiconductors.

Chinese companies have responded by improving domestic chips and linking larger numbers of processors together. A cluster can compensate partly for weaker individual chips when the system moves data efficiently and coordinates workloads across thousands of processors.

Huawei Technologies designed the Atlas 950 SuperPoD to connect large numbers of Ascend processors so they function as a unified computing platform.

The system does not prove that China has eliminated the technological gap with the United States. Performance depends on power use, software, memory, interconnect speed, reliability and the ability to manufacture processors at scale.

Its geopolitical importance lies in resilience. A functioning domestic computing ecosystem reduces the ability of export controls to halt Chinese artificial intelligence development completely.

China can also offer the system to overseas partners that face cost, availability or political barriers when purchasing United States technology.

That possibility adds an infrastructure layer to Beijing’s diplomatic proposal. Open models become more useful when countries can obtain compatible hardware, cloud services and training.

The United States must now consider whether restrictions will slow Chinese development or encourage the creation of a separate technology ecosystem less dependent on American suppliers.

The result may be a more divided global market in which governments choose between competing chips, software standards and governance frameworks.

What artificial intelligence safety principles did Xi Jinping introduce in Shanghai?

The July 17 address included some of the Chinese leader’s clearest public remarks on the risks created by advanced artificial intelligence.

He called for systems to remain under human control and supported early-warning and emergency-response mechanisms designed to identify emerging dangers.

The speech also addressed the possibility that autonomous systems could evade effective oversight. That concern places China within a growing international discussion about loss of control, model behaviour and the consequences of delegating decisions to increasingly capable machines.

Safety language serves several purposes for Beijing. It reassures governments that China is not promoting unrestricted development and gives the country a role in discussions involving frontier risks.

It also supports China’s preference for strong state authority over technology companies. The Chinese government can argue that public oversight is necessary to prevent private developers from creating uncontrolled systems.

The challenge will be translating general principles into verifiable rules. Governments differ over how dangerous systems should be tested, who should receive evaluation results and whether companies must disclose model capabilities.

Countries also disagree about military artificial intelligence. Human control over autonomous weapons and decision-support systems remains an unresolved international security issue.

China has called for cooperation while continuing to develop military and surveillance applications. The United States and other powers are doing the same.

Meaningful safety coordination will therefore require transparency among countries that view one another as strategic competitors.

The planned government-level artificial intelligence talks between Washington and Beijing could identify limited areas of common interest, including accident prevention, cybersecurity and control over autonomous systems.

Can China create a global artificial intelligence system without United States participation?

China can build an influential coalition, but a genuinely global governance framework would be difficult to establish without the United States.

American companies remain leaders in advanced chips, cloud computing, model development and artificial intelligence research. United States policy therefore affects technology supply chains, investment and safety standards far beyond its borders.

China has different advantages. It possesses a large domestic market, extensive manufacturing capacity, major technology companies and established political relationships across the Global South.

The two countries could create competing systems rather than a single international order. One might centre on American chips, proprietary models and alliances among advanced economies. The other could rely more heavily on Chinese infrastructure, open models and development partnerships.

Such fragmentation would create costs. Companies might need to comply with conflicting standards, while governments could face pressure to adopt one ecosystem.

Data might become harder to move across borders, and safety cooperation could weaken if the leading powers refuse to share information.

The United Nations could provide a broader forum, but geopolitical rivalry has limited consensus in other areas of international security.

The Shanghai proposal is therefore both an invitation and a challenge. China is offering cooperation while building institutions that can operate without Washington.

Whether that strategy produces a more inclusive system or a divided artificial intelligence order will depend on membership, transparency and the willingness of both powers to negotiate shared safeguards.

What are the key takeaways from Xi Jinping’s July 17 artificial intelligence speech?

• Xi Jinping used the opening of the World Artificial Intelligence Conference in Shanghai to present China as a leader of an open and cooperative international system that gives developing countries greater access to technology.
• China will provide 5,000 artificial intelligence training opportunities over five years and give 30 developing countries access to a Chinese meteorological system designed to improve forecasting and early-warning capabilities.
• Twenty-nine countries signed an agreement to establish the World Artificial Intelligence Cooperation Organization, a proposed intergovernmental institution headquartered in Shanghai and focused on cooperation, development and global governance.
• China plans to expand artificial intelligence partnerships through the Association of Southeast Asian Nations, African Union, League of Arab States, BRICS and other regional organisations where Beijing already maintains substantial diplomatic influence.
• Open-source and open-weight models are central to China’s strategy because they offer governments and developers lower-cost alternatives that can be adapted for local languages, industries and public services.
• Huawei Technologies showcased the Atlas 950 SuperPoD as evidence that Chinese companies are building large computing platforms around domestic processors despite United States restrictions on access to advanced semiconductors.
• The Chinese president called for human control, early-warning systems and emergency mechanisms to address artificial intelligence risks, including the possibility that autonomous systems could escape effective oversight.
• China can build an influential governance coalition without the United States, but competing American and Chinese technology ecosystems could fragment standards, supply chains, data rules and international safety cooperation.