China is seeking to position itself at the centre of the emerging international system for governing artificial intelligence, combining calls for global ethical standards with a broader strategy to strengthen its technological influence. A comprehensive action plan unveiled in Shanghai sets out proposals for international cooperation on AI safety, data protection, algorithmic accountability and inclusive development at a time when Chinese models are placing growing competitive pressure on established US technology companies.
The initiative reflects the increasingly close relationship between technological capability, regulation and geopolitical power. China is no longer presenting itself simply as a major developer or adopter of artificial intelligence. It is also attempting to shape the principles, technical standards and institutions that will determine how AI is tested, deployed and supervised worldwide.
China Proposes Global Guardrails for Artificial Intelligence
The Shanghai action plan moves beyond broad declarations about responsible innovation by outlining practical mechanisms for managing the risks associated with increasingly powerful AI systems. Its stated aim is to balance technological progress with protections against discrimination, privacy violations, unsafe deployment and the concentration of technological power.
Among the central proposals are international benchmarks for algorithmic accountability and stronger requirements for transparency. Developers would be expected to demonstrate that high-impact systems operate within clearly defined legal and social parameters, particularly when AI is used in sensitive sectors such as finance, healthcare, education, employment and criminal justice.
The plan gives particular attention to data governance. Modern AI systems depend on enormous quantities of information, much of which may include personal, commercially sensitive or copyrighted material. China’s proposed framework calls for stronger security, consent and oversight procedures governing the collection and use of data, including information transferred across national borders.
Algorithmic bias is another major focus. The plan supports audits designed to identify whether systems produce unfair outcomes for particular demographic or social groups. Such assessments could become a prerequisite before high-risk technologies are introduced into public services or commercial markets.
Technology companies are also encouraged to adopt an “ethics by design” approach. Rather than treating safety as a problem to be addressed after a product has been launched, developers would be required to identify risks during model design, training and testing. This would shift responsibility towards the organisations building AI systems and potentially require more detailed documentation of how models reach decisions.
The proposed governance structure depends on cooperation among governments, companies, universities and international organisations. Chinese officials have argued that national regulation alone cannot adequately address technologies that operate across borders. Shared incident-reporting systems, technical exchanges and international regulatory sandboxes could allow authorities to test emerging applications under controlled conditions before approving wider deployment.
Cooperation Becomes an Instrument of Technological Influence
China’s emphasis on multilateral cooperation also carries significant strategic implications. By helping to establish global standards, Beijing could influence the operating conditions faced by companies and governments throughout the international AI economy.
The initiative presents universal ethical principles as a basis for cooperation between countries with different political systems and economic interests. China has placed particular emphasis on involving developing nations, arguing that AI governance should not be dictated exclusively by a small number of wealthy countries or Western technology corporations.
This approach allows Beijing to present itself as an advocate of a more inclusive technological order. Shared research programmes could address problems such as model interpretability, cultural bias and the development of multilingual systems for regions that remain underrepresented in existing datasets.
Cross-border data governance is likely to remain one of the most difficult areas. Governments want to benefit from international research and model training while protecting national security, personal information and commercially valuable datasets. China’s proposals seek to create a more predictable framework for these exchanges, although differences over surveillance, censorship and state access to data may complicate negotiations.
The focus on cooperation therefore serves several purposes. It may reduce some of the geopolitical friction surrounding Chinese technology, provide Chinese companies with greater access to overseas markets and allow Beijing to influence international rules before those rules become firmly established elsewhere.
Global standards can also have commercial consequences. Once a governance framework becomes widely accepted, companies may need to comply with its audit, reporting and safety requirements to enter participating markets. Chinese-aligned standards could therefore shape product development well beyond China’s borders, just as European privacy and digital regulations have influenced the practices of global companies.
Chinese AI Advances Reshape Global Market Expectations
China’s diplomatic push comes as its AI companies are demonstrating models capable of competing more directly with leading US systems. These advances have contributed to volatility in technology markets, with investors reassessing assumptions that American companies will maintain an uncontested lead in AI hardware, software and infrastructure.
Nvidia and other major US technology stocks have faced periods of selling pressure following evidence that Chinese developers can produce increasingly capable models using fewer resources or alternative computing strategies. The concern is not that demand for advanced chips will disappear, but that the global market may become more fragmented and competitive than earlier forecasts suggested.
Chinese companies are investing in domestic processors, customised computing systems and proprietary software platforms. This vertical integration could gradually reduce dependence on foreign suppliers while allowing developers to optimise models for locally available hardware. Export restrictions imposed by Washington have accelerated the search for alternatives, encouraging Chinese firms to improve efficiency and build more resilient supply chains.
US companies still retain major advantages in advanced semiconductor design, cloud infrastructure and access to global developer communities. Nevertheless, rapid progress by Chinese laboratories has challenged market valuations based on the assumption of prolonged Western dominance.
The emerging contest is no longer concerned solely with which country can train the largest model or deploy the fastest processors. It increasingly involves the ability to create a complete AI ecosystem encompassing chips, energy, data, research talent, commercial applications, safety standards and international partnerships.
China’s Shanghai plan brings these strands together. It presents ethical governance and international cooperation not as separate from technological competition, but as essential components of it. By advancing domestic capabilities while promoting global rules, Beijing is attempting to become both a leading producer of artificial intelligence and an architect of the system within which the technology develops.
