China's Pursuit of "Open AI Supremacy" — A Future Different from the U.S. as Envisioned by Xi Jinping

The future of artificial intelligence is not something that only the giant tech companies in the United States will decide.

The message delivered by Chinese President Xi Jinping in Shanghai can be summarized as such.

On July 17, 2026, at the opening ceremony of the World Artificial Intelligence Conference, Xi argued that the development of AI should not be a "solo performance by one country" but rather a "symphony of global cooperation."

He did not specifically name the United States in his speech. However, it was clear where his criticism was directed.

In response to U.S.-led frameworks such as export restrictions on advanced semiconductors, security policies surrounding AI models, and closed technology development centered around American companies, China is presenting another AI world order with words like "openness," "sharing," and "support for developing countries."

This is not merely a technological policy. It is a diplomatic strategy centered around AI and a struggle for leadership over international standards.

China Advocates "Openness"

Xi particularly emphasized the potential of open-source AI.

The mechanism of making software and models available to the public so that developers and companies can improve, verify, and apply them is considered an important means of preventing AI capabilities from being concentrated in a few companies.

In China, companies like DeepSeek, Moonshot AI, and Zhipu AI are successively launching open models. The performance gap with cutting-edge models provided by American companies is narrowing, and in terms of cost and ease of implementation, Chinese models are increasingly becoming an attractive option.

However, caution is needed with the term "open source." Even if the model's weights are public, it does not necessarily mean that all the training data, development process, and training code are disclosed. In many cases, it is more accurate to call them "open weight" models.

Nonetheless, the significance of companies and research institutions being able to run them on their own servers and adjust them according to their needs is substantial. It allows them to implement AI without relying on American companies' cloud services.

For China, open models are also a means to compensate for its disadvantage in technological competition with the United States.

Instead of enclosing it as a single company's product, they aim to have developers and startups worldwide use it. As the number of users increases, there is a higher possibility that Chinese-originated technical specifications and development methods will become de facto standards.

In other words, China's "openness" is both a charitable act and a market expansion strategy.

AI Diplomacy Toward Developing Countries

A key point in Xi's speech was positioning the access gap to AI as an issue leading to "historical injustice."

The gap between countries that can use AI and those that cannot is not just a digital divide. It has the potential to influence productivity and national capabilities in all fields, including manufacturing, healthcare, education, agriculture, administration, disaster prevention, and the military.

If only developed countries possess high-performance AI and massive computing resources, existing economic disparities will widen further.

China is trying to position itself as a "partner providing accessible AI" to emerging and developing countries that harbor these concerns.

Xi announced plans to offer 5,000 AI training and seminar opportunities to developing countries over the next five years. He also indicated plans to establish cooperation bases with ASEAN, the African Union, Arab countries, Latin American countries, and BRICS.

If Chinese AI is introduced in fields such as early warning for weather disasters, agricultural production forecasting, medical diagnosis, and administrative services, China will become deeply involved in the digital infrastructure of partner countries.

In the past, China expanded its influence overseas through ports, railways, power plants, and communication networks. In the future, AI models, data centers, cloud services, and technical education may become the new center of infrastructure diplomacy.

Following the "Belt and Road" initiative of developing roads and ports, a "Digital Belt and Road" based on AI is also forming.

An AI Order Different from the U.S.

In the current cutting-edge AI market, American companies hold a significant presence. OpenAI, Anthropic, and Google develop high-performance models and provide services to companies and consumers worldwide.

Meanwhile, the most capable models are generally managed internally by companies. Users access the models via APIs or web services, but much of the internal structure and training methods are not disclosed.

This approach has advantages in terms of safety management, intellectual property protection, and monetization.

However, there are also issues such as companies setting usage fees and conditions, being affected by service suspensions or changes in terms, and users not being able to fully manage the models.

China criticizes this structure as a "monopoly by some countries and companies" and presents its open models as a counter-axis.

In Shanghai, an international AI cooperation organization led by China also showed its presence. This framework, involving multiple countries, serves as a diplomatic foundation to reflect China's intentions in AI rule-making, safety standards, talent development, and technical cooperation.

The U.S. is also building a cooperative framework with friendly countries regarding semiconductors, data centers, critical minerals, and AI supply chains.

The global AI order may not consolidate into a single common market but instead divide into a U.S.-centered camp and a China-centered camp.

However, many countries do not want to choose just one side. Some countries may make pragmatic choices, using American semiconductors and cloud services while also adopting low-cost Chinese models.

International relations surrounding AI are likely to become a complex structure of utilizing multiple technological spheres rather than a simple bipolar confrontation.

The Reason for Emphasizing "Human Management"

Xi mentioned not only openness but also the safety of AI.

He argued for the need to place AI under human management and establish systems for monitoring, early warning, and emergency response. He also touched on the risk of autonomous AI deviating from human intentions and becoming uncontrollable.

This indicates that China is not unconditionally promoting the rapid development of AI.

At the same time, the expression "managed by humans" has a different meaning in China's political system. It raises questions about who manages AI and which information is permitted or prohibited.

Generative AI provided domestically in China is subject to strict content management. The government has strong influence over politically sensitive topics, handling of training data, and the responsibilities of service providers.

Therefore, some people find it contradictory for China to advocate "open AI" to the international community.

While open to the use and improvement of technology, it does not necessarily mean openness to information freedom or political expression.

The openness proposed by China is not the same as the liberal openness discussed in the West. It is closer to the idea of sharing technology and economic activities while the state retains control over social and political authority.

Voices of Expectation on Social Media

In online communities where overseas AI technologists gather, posts evaluating China's open model strategy stood out regarding the recent speech.

Especially among developers running large language models in local environments, there was a view that "if China continues to invest in open AI, the dominance of American companies offering closed models will crumble."

There was also analysis suggesting that, just as Chinese companies made massive investments and reduced costs in solar panels and electric vehicles, they might be trying to turn AI models into low-cost products.

With more open models, researchers, small businesses, and individual developers can use advanced AI without paying high fees to large companies.

Some responses appreciated that "China has many issues, but its clear stance on technical support for developing countries is more realistic than American AI companies."

For these users, what matters is not the political system of the country that developed the model, but the price, performance, modifiability, and freedom of local execution.

If Chinese AI is sufficiently high-performance, inexpensive, and can be modified according to needs, there are few reasons to hesitate in adopting it.

Voices Viewing It as "Not Benevolence but National Strategy"

On the other hand, many pointed out that Xi's speech should not be taken as mere idealism.

On social media, there was a view that "China's promotion of open AI is not solely for the benefit of humanity but to prevent power from concentrating in a single company and ultimately maintain state control."

For the Chinese government, it is undesirable for domestic giant AI companies to hold more influence than the state. The policy of making technology public and encouraging competition among multiple companies prevents monopolies while maintaining government control over companies.

Moreover, if Chinese models penetrate worldwide administration, communication, finance, and healthcare, user countries may become dependent on China's technological infrastructure.

Even if the model itself can be run on a country's own servers, dependence on Chinese companies for updates, developer communities, computing equipment, maintenance services, and training data could create a new form of enclosure.

"Being able to use it for free or cheaply" and "being able to be self-sufficient in the long term" are not the same.

Concerns about data protection and supply chains were also expressed by users sensitive to security. There is anxiety that incorporating Chinese AI into critical infrastructure could lead to influence through future updates or vulnerabilities.

Emphasis on "Public Goods" in the Chinese-speaking World

On Chinese social media, posts introducing Xi's speech positively, mainly from state media and policy commentary accounts, were frequent.

The concept of positioning AI as an international public good that benefits all humanity and comprehensively sharing data, computing resources, open models, talent development, international standards, and safety management was emphasized.

Policy-related posts also introduced the practical aspect of introducing AI into manufacturing, healthcare, education, agriculture, disaster prevention, and administration to solve social issues in developing countries.

However, due to restrictions on political expression in China's social media space, it is not possible to determine that the entire Chinese society supports the policy just by looking at the displayed reactions.

On the other hand, English-speaking social media like Reddit also have strong political tendencies depending on the user base and community. The posts confirmed this time are not opinion polls but examples indicating the trend of discussions.

The Real Issue is Not Just "Performance"

So far, the U.S.-China AI competition has often been discussed from the perspective of which model has higher performance.

However, Xi's speech indicated the reality that the scope of competition has greatly expanded.

In the future, the competition will not be only about model performance.

Who will supply semiconductors? Who will build data centers? Which country's technical standards will be adopted? Which model will developers use as a basis for creating apps? Which country will train technicians in developing countries? And who will decide the international rules regulating AI?

China is trying to capture global users first with open models and low prices, rather than waiting to completely surpass the U.S.

In that sense, open source is both a technical philosophy and a geopolitical weapon.

As American companies continue their strategy centered on high-performance closed models, if China increases its presence as a supplier of open models, global developers will have to choose between two different AI ecosystems.

Will China's proposal truly support the independence of developing countries, or will it create new dependencies on China?

How will the international community evaluate the contradiction of advocating openness while continuing information control domestically?

The answer is not yet clear.

However, one thing is certain. China no longer accepts the position of merely following the AI order created by the U.S.

The "symphony of global cooperation" spoken of by Xi Jinping in Shanghai was both a call for collaboration and a declaration of the start of competition over who will be the conductor.

Sources and References

New York Times: Article reporting Xi Jinping's speech positioning China as a leader in open AI development and support for developing countries.
https://www.nytimes.com/2026/07/17/business/xi-jinping-china-ai.html

Xinhua News Agency: Full text of Xi Jinping's keynote speech at the 2026 World Artificial Intelligence Conference. Used for confirming official expressions and policy directions.
https://english.news.cn/20260717/893fe11097db460ea31b98f131e34ef0/c.html

Reuters: Article reporting on China's AI cooperation organization, support for developing countries, competition with U.S.-led frameworks, and the expansion of Chinese open models.
https://www.reuters.com/world/asia-pacific/chinas-xi-promotes-chinas-commitment-ai-access-speech-shanghai-conference-2026-07-17/

Associated Press: Article reporting specific measures from the speech, including 5,000 AI training sessions, provision of weather technology to 30 countries, Chinese AI models, and international cooperation organizations.
https://apnews.com/article/china-ai-tech-chips-xi-us-df4cfc7e1b260e765b5449b6d71a48e5

Reddit・LocalLLaMA: Confirmation of reactions from the tech community regarding the potential threat of large-scale investment in open AI to U.S. closed AI companies.
https://www.reddit.com/r/LocalLLaMA/comments/1uytamy/chinese_president_xi_jinping_speaks_at_world_ai/

Reddit・Singularity: Examples of posts positively evaluating support for developing countries and