Xi Jinping's Big AI Speech, Annotated

My first-pass analysis of his keynote at Shanghai's World AI Conference

Hey there,

Today I’m coming to you with my first reactions to Xi Jinping’s much-hyped keynote address to the World AI Conference (WAIC) in Shanghai. WAIC has turned into China’s biggest AI event of the year, and 2026 is the first time that Xi Jinping has come in person. That’s a significant signal and this is a very significant speech.

Speeches by Chinese officials can come off as very boring — and many of them are! But the words are always chosen very carefully and some of them are worth parsing closely. This is one of those. The speech is Xi’s most in-depth discussion of AI to date, and its coming at an incredibly dynamic moment in AI development and policy.

Up top here I will give three of my key takeaways from the speech, and below I have included the full English translation of the speech, with my commentary on specific pieces of it included as footnotes. (29 in total!) This is a new format for me so we’ll see how it works — let me know how the footnotes / annotations worked for you.

Further down I’ve included a suggestion on how best to read the annotated speech. Tldr: Read it in a browser (just click on this article’s title at the top of this email) or in the Substack app.

Three takeaways from the speech:

1. Xi doubled-down on China’s pitch to the world on AI: We are the open, cooperative partner who wants to help you build your own capabilities.

This has been the pitch for 3+ years and it’s implicitly (though very clearly) framed as a contrast to the United States. Xi’s 2023 Global AI Governance Initiative called out those unnamed countries for “forming exclusive groups to obstruct other countries from developing AI.” Since then, China has made AI capacity building for the Global South core to its pitch, culminating in the founding of the World AI Cooperation Organization (WAICO) in Shanghai this week. How successful that pitch and how impactful the capacity building ends up being is an open question.

2. Xi restated the status-quo on open-source: we should “encourage open-source AI.”

Many people in international AI policy circles were watching this speech for any signal on whether China is likely to walk back or hedge it’s support for Chinese companies releasing their models open-weight on the internet. The speech didn’t signal that change. Restating the status-quo is definitely notable, but I don’t take this as a major update in either direction. OS was mentioned in just that one line in the speech and not as a stark pledge about future policy directions. Also…

3. Xi concluded the speech with stark language on remaining ready to change course as circumstances change, and preventing “loss of control.”

Chinese official discussions of AI often include references to “agile governance” and keeping abreast of new trends in the technology. But the language used here was significantly stronger than normal. It was then followed by a rich sentence on how the “staggering speed” of AI development requires, among other things, that we “constantly refine measures to forestall loss of control.” As I note below, in Chinese AI policy discourse “loss of control” is used to refer to different things in different contexts: state/sovereign control, human control, industrial control, etc. More on that in the annotations below.

---

The words Xi uses are significant, both because of what they signal about his preferences and because people across the system will leverage the fact that he said them to advance their own goals. In everything from regulatory fights to major grant applications, being able to tie your thing directly to Xi’s words helps. But many of those words remain open to interpretation. I’ll be closely following how they get interpreted—and, more importantly, acted upon—in the year ahead.

Please remember: There’s a lot to digest from this one, and these are my first impressions of the speech without reading in-depth commentary from others. There are almost certainly things that I overstated / missed / misunderstood in here, so take it as a first draft analysis. I hope it’s helpful and look forward to reading what others think.

Note: I recommend reading the annotations either in a web browser or in the app. That way you can hover over a footnote number or click on it. If you hover over it, the footnote’s text will appear as a pop-up. If you click on a footnote, it will take you down to the bottom and you can click on that footnote number at the bottom to return to original spot in the body text.

If you read in email form, I believe you have to scroll up and down yourself.

The footnotes in the text appear small, so I’ve highlighted each one to make it easier to scan for them. I think the annotations get more interesting as the speech goes along, so stick with it or scroll down.

You can find the original Chinese text of the speech here.

Xi’s Full Speech at WAIC 2026

[Before Xi’s speech, Shanghai Party Secretary Chen Jining gave welcome remarks.¹]

Distinguished Colleagues and Guests,
Ladies and Gentlemen,
Friends,

Seventy years ago, a group of young scholars proposed the concept of artificial intelligence (AI) for the first time at the Dartmouth Workshop in New Hampshire of the United States². In the subsequent 70 years, AI scientists and researchers from around the world ventured into this unknown territory, forged ahead through twists and turns, and made breakthroughs with persistent hard work. Seven decades later today, amid the new wave of AI development, we are gathering by the Huangpu River to discuss how to promote AI globally for the positive, for good and for humanity. All this makes our meeting highly important. On behalf of the Chinese government and people, I would like to extend a warm welcome to you all.

In the course of history, the invention of the steam engine heralded the industrial civilization, the widespread access to electricity brightened up modern society, and the birth of the Internet brought the entire world together³. Each of these technological revolutions has profoundly reshaped our way of work and life, and enabled a giant leap in economic and social development.

Today, major changes unseen in a century are accelerating across the world⁴. The new round of technological revolution and industrial transformation is advancing at a faster pace, and the world has entered an unprecedented period of active innovation on AI technologies. Intelligent connectivity, human-machine collaboration, cross-sector integration, joint creation and sharing, and other intelligent technologies are unleashing enormous power. All this carries within it great opportunities as well as challenges to governance. We human beings must answer the questions posed by our times: How to get along with thinking machines? How to ensure security⁵ when algorithm is part of decision making? How to tackle ethical challenges by technologies through adaptive governance?⁶ How to realize AI for all when the divide keeps widening? These questions demand serious consideration and real answers from the whole international community.

In China’s view, all countries should take a people-centered approach⁷ and develop AI for the positive and for good. We should ensure that AI is an important driver for shared prosperity and common security. We should join hands to build a just and equitable system for global AI governance. To this end, I wish to share four observations.

First, we should adhere to the principle of openness⁸ and win-win and boost innovation-driven development. As a new engine of world economic growth and an accelerator for the shift of growth drivers, AI is moving from the digital world into the physical world. We should seize this rare, historic opportunity to encourage open source⁹, openness, collaboration and sharing. We should facilitate technological innovation, industrial development and scenario-based application of AI. We should make coordinated advances in the transformation and upgrade of traditional industries¹⁰, the cultivation and growth of emerging industries and forward-looking planning for future industries, so that all sectors and businesses can benefit from AI.

Second, we should strengthen risk-awareness and ensure that AI is secure and controllable¹¹. AI should be a trusted tool for humanity. We should take seriously the various types of inherent and secondary risks¹² that AI may trigger. We should put in place laws and regulations, technological monitoring, early warning and emergency response systems in order to strengthen the line of security¹³, prevent abuses and malicious use, and ensure that AI is always under human control¹⁴. In the meantime, we should jointly oppose overstretching the national security concept in the field of AI and placing one country’s security over that of others.¹⁵

Third, we should encourage inclusiveness and promote mutual learning between civilizations. AI development and its application should not erode or undermine the diversity of world civilizations or the uniqueness of cultures of different countries.¹⁶ We must shape the values of AI with humanity’s common values, and make good use of AI technologies to increase understanding, tolerance, exchanges, and sharing among all civilizations. We should tend to the garden of civilizations with great care to ensure that the beauty of each civilization is appreciated and shared.

Fourth, we should advocate solidarity and improve global governance. AI is an invaluable asset that encapsulates humanity’s collective wisdom. We should practice true multilateralism and recognize the important role of the United Nations.¹⁷ We should enhance alignment and coordination on AI development strategies, governance rules and technical standards, so as to form a consensus-based global governance framework at an early date to make this frontier technology better benefit humanity. We must carry out extensive international cooperation and help Global South countries with capacity building to bridge the AI and digital divides, promote sustainable development, and prevent creating new historical injustice in AI.¹⁸

Ladies and Gentlemen,
Friends,

This year marks the start of China’s 15th Five-Year Plan. It maps out China’s economic and social development for the next five years and provides immense opportunities for the international community. In recent years, China has embraced AI with open arms. We have promoted interplay between an efficient market and a well-functioning government, strengthened AI innovation¹⁹, actively advanced the AI Plus Initiative,²⁰ and built a healthy ecosystem for all entities to thrive in together. The core smart economy industries are worth at least RMB 1 trillion yuan.²¹ Smart devices in countless homes truly improve people’s livelihood. Intelligent Manufacturing in China has become another shining hallmark of Chinese modernization.

At the same time, China lays great emphasis on safety and security in AI development.²² With a deep understanding of the trends and logic of AI development, we are continuously improving laws, regulations, policies, mechanisms, application norms, as well as ethical principles to make sure that AI is safe, secure and controllable, and that this fine steed of AI gallops with both speed and stability.

As a responsible major country, China is always committed to providing international public goods relating to AI. Since I proposed the Global AI Governance Initiative, China has promoted the adoption of the U.N. General Assembly Resolution on Enhancing International Cooperation on Capacity Building of Artificial Intelligence by consensus, published the AI Capacity Building Action Plan for Good and for All, announced the AI Plus International Cooperation Initiative, and advocated for establishing the World Artificial Intelligence Cooperation Organization (WAICO). China has been contributing steadily to the global AI governance.²³

We often say in China, “A single string cannot make music, and a single tree does not make a forest.” AI development should not be a solo performance by a single country, but a symphony of international cooperation.²⁴ Thanks to our joint efforts, WAICO has come into being in Shanghai. Our vision from one year ago is now a reality. This is a major move by China to answer the call of the Global South and unite the international community together to promote vigorously AI development and governance. It will be an important milestone in the history of AI development.

To further support global AI development and advance global AI capacity building, I hereby announce that in the next five years, China will provide developing countries with 5,000 opportunities in AI training and seminar programs²⁵; develop international AI application cooperation centers with the Association of Southeast Asian Nations, the League of Arab States, the African Union, the Community of Latin American and Caribbean States, the Shanghai Cooperation Organization, and BRICS; and enable 30 countries to use the AI-powered meteorological warning system, or MAZU,²⁶ to safeguard homes around the world.

Ladies and Gentlemen,
Friends,

As ancient Chinese observed, “A man of wisdom adapts to changes; a man of knowledge acts by circumstances.”²⁷ With AI advancing at a staggering speed, we must ensure its development is for the positive, for good and for humanity. We must make its oversight and governance precise and effective, and constantly refine measures to forestall loss of control.²⁸ We should always guide AI development with human wisdom and international consensus, so that AI can truly become a mighty force that increases the well-being of humanity and advances human civilization.²⁹

China is ready to be more open, take more practical actions, and assume a more visionary perspective. We are ready to work with all parties to seize the opportunities of AI development and meet the challenges, and join hands to create a brighter future for humanity.

Thank you!

---

Matt here again —

I’ll leave it there. Again, consider this a first draft analysis, and one that’s very open to change. Would love to hear others’ take on the speech and feedback on this.

Enjoy your weekend!

1. Before jumping into the speech, a note on Chen Jining. It’s not surprising that he gave the welcome remarks — WAIC is Shanghai’s baby. But Chen seems to be very interested and involved in AI beyond WAIC, and I wouldn’t be surprised if he took on a major role in AI policy down the road. I’m not a deep CCP elite politics watcher—just my take.

2. Must say, I was a bit surprised to see Xi mention the U.S. in the first sentence of the speech, crediting it as the birthplace of the AI field. I wouldn’t call it an olive branch but if the relationship was more openly confrontational right now I doubt he would have included this.

3. Historical aside: The World AI Conference (WAIC) is loosely modeled on China’s World Internet Conference (WIC), which launched in 2014 in Wuzhen. It was a moment of burgeoning confidence about China’s booming internet economy. China was transitioning from not wanting to talk about its internet controls to instead proudly pitching them globally as “Internet sovereignty” — essentially, every state having the right to control its domestic internet as it pleases.

The pitch at the first conference was clumsy. In 2014, the WIC organizers shoved a piece of paper with the “Wuzhen Declaration” under people’s hotel doors in the middle of the night, and told them that if they didn’t officially object by morning then they would be considered to have co-signed it. The foreign participants vehemently objected and the plan was shelved. WAIC has been much more smooth and sophisticated with this kind of thing, though they have had a number of years of practice.

Worth noting: WIC has been the CAC’s baby since the beginning, while WAIC is led by Shanghai city, with the MFA handling international aspects. There’s an element of bureaucratic competition there that I believe affects how different ministries engage with WAIC.

4. A standard phrase (世界百年变局) that Xi regularly uses. It’s often read as meaning the seismic change of a superpower transition (aka China surpassing the U.S.), though official Chinese sources would present it differently.

5. Mildly interesting that the official English goes with “security” here instead of “safety.” The Chinese word (安全) can be translated as either one, depending on context and intent. Chinese AI policy documents toggle between the two translations, not always in a super logical way.

6. More direct translation would be “When technology challenges ethics, how can governance keep up?”

7. The “people-centered” or “human-centered” approach is a long term theme in Chinese policy, not new to AI.

8. Impt: “Openness” does not always equate to “open-source.” China frequently invokes “openness” when pitching all kinds of things to the world, especially trade and investment.

9. This line is the clear and direct endorsement of open-source in the speech. Right now there is tons of debate in western AI policy circles about whether China might shift it’s position on allowing the open-sourcing of its most frontier models. I take this line in the speech to be a restatement of China’s long-standing language on this — the 2023 Global AI Governance initiative included a very similar single line that mentioned “encouraging” open-source AI.

Now, you could say that restating the standard line constitutes a signal that they will continue on the OS pathway. Yes, to an extent (as in, not going to close up tomorrow). But before hearing the speech, I had expected more or stronger endorsements if they were 100% set on this pathway. OS models are China’s biggest selling point on the global stage and large parts of this speech were direct pitches to those other countries. What gives me additional pause on this are two sentences in the second-to-last paragraph on the “staggering” speed of advancement and “adapting to change.” More on that in Footnote 26.

10. This focus on the upgrading / efficiency-maxxing of traditional industries is an absolutely essential part of China’s vision for AI diffusion and (information) technology more broadly.

11. Ensuring AI is “controllable” is one of the longest running AI policy terms in China. What they mean by “controllable” has evolved over time and shouldn’t always be read as referring to the “loss of human control” concern prominent in AI safety circles. In the 2010s they used “controllable” to describe a vague idea of sovereign/state control over the AI stack. Beginning in the early 2020s we saw explicit, macro-level “loss of human control,” and nowadays it is also invoked in an “industrial control” context.

12. “Inherent and secondary risks” is a relatively new formulation that I think first came to prominence with the 2025 AI Safety Governance Framework 2.0 (my + Scott Singer’s analysis here). Here is how that document defined the difference:

“AI entails not only inherent technical risks such as flaws in models and algorithms and the poor quality of training data and corpora, but also application-level risks in areas such as network systems and information and content during technology integration and deployment. Risks could also arise from misuse, abuse, and malicious use of technology, resulting in real-world and cognitive [secondary] risks, and even catastrophic risks.”
^^Note on a note: the Chinese version included the “secondary risks” term, but not the official English translation.

13. Lots packed in here. “Technological monitoring, early warning, and emergency response systems” have been invoked regularly over the past year or two. We don’t yet know how these might be instantiated, but I would caution against AI safety-focused people reading too much into some of these terms. China’s first technical document on “emergency response” for generative AI models was primarily focused on “content risks,” with some on data and cybersecurity and model tampering as well.

That said, the “line of security” (安全底线) part is a notable phrase. A more direct translation would be the “safety bottom-line,” and maintaining “bottom-line thinking” is a core theme in discussions of national security and other safety-related issues.

14. This is the standard, explicit formulation used when referring to macro-level “human control” over AI (确保人工智能始终处于人类控制之下, could also be translated as “under humanity’s control”). This specific formulation first appeared (I believe) in the 2021 “Ethical Norms for New Generation AI” document, which was published by the Ministry of Science and Technology’s AI governance expert advisory committee (chaired by Dean Xue Lan of Tsinghua.) Two years later, it showed up in the Global AI Governance Initiative, receiving Xi’s personal imprimatur. Since then it’s continued to appear in documents.

15. A not too subtle dig at the US.

16. This language on AI steamrolling cultural diversity often falls on deaf ears in the DC policy world, but I think this is something that people in lots of countries (especially small ones, especially the Global South) see as a very real concern.

17. China always pushes for the UN to play the central role in international governance. In the announcement of the new World AI Cooperation Organization, China took great pains to say that it is aligned with / supportive of the UN’s role, not trying to replace it.

18. Quite pointed line on “historical injustices.” China has it’s own master-narrative about the historical injustice it suffered (the “Century of Humiliation”), caused in large part because it had missed out on the industrial revolution, and thus was weak and easily victimized by foreign powers.

19. This is China’s core pitch to the international business community.

20. For a deep dive on AI+, check out: China’s Big AI Diffusion Plan is Here. Will it Work?

21. China’s 2017 national AI plan set a goal of AI’s core industry reaching 400 billion RMB ($60b USD) by 2025. This claims China has more than doubled that (by 2026).

22. The Chinese version uses a standard formulation invoked in many AI policy documents: China always places “equal emphasis on safety and security.” (发展与安全并重)

23. China has made AI capacity building for the Global South core to its pitch on international AI policy, and it rolled out the documents in the above paragraph covering this. We don’t yet know what level of real-world action will emerge out of these documents. Personally, my suspicion is that there will be less follow through here than on things like the Belt and Road Initiative, mainly due to the downturn in the economy and general drying up of government revenue. But I could be wrong.

24. Not a bad visual metaphor for what China is arguing.

25. The Chinese term here (名额) seems to indicates this will be 5,000 people receiving training, not that there will be 5,000 seminars. (Please correct me if I’m wrong on this.) Depending on the nature of the training and how extensive it is this could end up being a very modest program — or a big one if we’re talking about training PhDs or something, but that’s not how I read it.

26. The subject of a big announcement at a previous WAIC. I haven’t learned much about it but would be worth reading up on it.

27. This is a striking choice of quotation to begin the conclusion of the speech. I could be over-reading this, but to me it suggests: “What I’ve said here today is the current plan. But if circumstances (or the technology) really change, we have to be ready to change our approach.” Not a bad mindset to have going into the next few years of AI development and policy.

28. Very rich sentence. The official translation isn’t fundamentally misleading but it’s worth looking at a more literal translation (via Claude):

"The faster artificial intelligence technology develops—advancing by leaps and bounds—the more precisely its direction must be anchored toward progress, goodness, and benefiting humanity; the more accurately the scale and measure of regulatory governance must be grasped; and the more promptly measures to prevent loss of control must be improved and refined."

I would again caution that when Xi says “loss of control” it does not always refer to “loss of human control” — for example, it’s been used to mean AI-powered industrial systems going awry. But this is a pretty striking choice for how to (basically) close-out the speech, especially when much of the speech is clearly pitched at developing countries and the international business community. To my understanding, technical and policy work on testing for / preventing loss of control is in a relatively early stage, but this kind of language could be a big boost to those working on these areas.

29. Nothing terribly knew in this sentence, but it reminded me of this fascinating essay that was published in Qiushi (the Party’s flagship ideological journal) this week: “Let Technology Forever Remain a Tool of Humans.” Bill Bishop shared a translation here, and I shared some very brief thoughts in this note.