On 17 July, Xi Jinping delivered the keynote address at the opening of the 2026 World AI Conference in Shanghai – his first in-person appearance at the conference since it began in 2018, and coming a day after 29 countries signed the agreement establishing the World Artificial Intelligence Cooperation Organisation (WAICO), a new intergovernmental body headquartered in Shanghai. The event was attended by President Kassym-Jomart Tokayev of the Republic of Kazakhstan, Prime Minister Hun Manet of the Kingdom of Cambodia, Prime Minister Anutin Charnvirakul of the Kingdom of Thailand, Secretary General of the United Nations António Guterres, and official representatives, business leaders, scholars and researchers from more than 100 countries and international organisations.
In the following analysis for Friends of Socialist China, Carlos Martinez examines the significance of Xi’s speech and the launch of WAICO. He argues that they represent a direct alternative to Washington’s approach – encapsulated in the Trump administration’s “Winning the Race” AI action plan, with its export controls, model restrictions and pressure on allies to join a technological blockade of China.
Where the United States treats artificial intelligence as a weapon in a zero-sum contest for supremacy, Xi calls for “open source, openness, collaboration and sharing”, and for AI to be developed as a common asset of humanity, with its benefits extended to the Global South. It is, Carlos writes, a democratic, multipolar and socialist vision: technology as a public good, serving people before profit.
This article was first published by Friends of Socialist China.
Should artificial intelligence – perhaps the most consequential technology of our era – be the exclusive property of a handful of corporations and one increasingly belligerent superpower, or should it be developed as a common asset of humanity? That is the question Xi Jinping posed, implicitly but unmistakably, in his keynote speech at the opening of the 2026 World AI Conference in Shanghai on 17 July – his first in-person address to the conference since its launch in 2018, and delivered a day after 29 countries signed the agreement establishing the World Artificial Intelligence Cooperation Organisation (WAICO), a new intergovernmental body headquartered in Shanghai.
Citing an old Chinese saying, Xi observed that “a single string cannot make music, and a single tree does not make a forest”, insisting that AI development “should not be a solo performance by a single country, but a symphony of international cooperation”. The metaphor is well chosen, because a solo performance is precisely what Washington has in mind.
Winning the race, or sharing the prize?
The United States announced its own AI strategy a year ago. Its title – Winning the Race: America’s AI Action Plan – tells us much of what we need to know about the underlying philosophy. AI is conceived not as a tool for human development but as a weapon in a zero-sum contest for global supremacy; the plan is built around tightening export controls on advanced semiconductors, restricting access to leading US models, and pressuring allies – under threat of secondary tariffs – to join a technological blockade of China. The Trump administration has since extended export controls to cutting-edge US AI models, forcing even European businesses to confront the risks of depending on US technology.
Xi’s speech reads as a point-by-point repudiation of this approach. Where Washington seeks to hoard the technology, China calls for “open source, openness, collaboration and sharing”. Where Washington weaponises “national security” to justify its chip war, Xi calls on the international community to “jointly oppose overstretching the national security concept in the field of AI and placing one country’s security over that of others”. Where Washington demands that the world adopt “American values” along with the American AI stack, Xi insists that AI “should not erode or undermine the diversity of world civilisations or the uniqueness of cultures of different countries”.
The West’s framing of AI as a civilisational contest is far from subtle. At the Paris AI summit last year, JD Vance warned darkly of “authoritarian regimes” using AI to control their citizens, while Britain’s technology secretary Peter Kyle declared that the AI race must be led by “western, liberal, democratic” countries. China’s response at the time – that it is “against drawing lines along ideological difference, overstretching the concept of national security, or politicising trade and tech issues” – is precisely the position Xi has now elevated to the level of a global governance programme. And as legal scholar Angela Zhang observed in the Financial Times, China does not think in terms of an “AI race”; its priority is not supremacy but self-sufficiency, diffusion and application – making the technology useful, cheap and universal.
People before profit
The contrast between China and the West’s approaches reflects two different social systems, with two different logics. In the US, AI development is driven by the profit expectations of a handful of tech giants and the war-planning of the Pentagon. In China, a socialist market economy directs the technology towards social ends: Xi spoke of smart devices that “truly improve people’s livelihood”, of coordinated upgrading across traditional and emerging industries “so that all sectors and businesses can benefit from AI”, and of keeping AI “always under human control”.
The difference shows up in practice. While Silicon Valley promises mass layoffs as a selling point to investors, China’s AI Plus programme – now fully integrated into state industrial policy – is oriented towards job creation and job quality, popularising AI as a social good while regulating its disruptions.
In Chinese hospitals, an AI screening tool called PANDA is detecting pancreatic cancer from routine CT scans before symptoms appear – catching early, at population scale, one of the deadliest cancers. And while Wall Street inflates an AI investment bubble of historic proportions, the CPC’s top theoretical journal has been urging the cultivation of “patient capital” – long-term, mission-oriented investment in place of speculative frenzy.
The DeepSeek phenomenon exemplifies the whole approach: a world-class model, developed at a fraction of the cost of its US competitors, requiring far less energy and computing power, released free and open source to the entire world. One model treats intelligence as a commodity to be enclosed; the other treats it as a public good to be shared.
Answering the call of the Global South
Some of the most striking passages of Xi’s speech concern the developing world. Xi warned against allowing the AI revolution to create a “new historical injustice” – a digital reprise of the colonial division of the world – and committed China to “help Global South countries with capacity building to bridge the AI and digital divides”. This was backed with concrete pledges: 5,000 AI training opportunities for developing countries over the next five years; international AI application cooperation centres with ASEAN, the Arab League, the African Union, CELAC, the SCO and BRICS; and the extension of MAZU, China’s AI-powered meteorological early warning system, to 30 countries.
WAICO itself is, in Xi’s words, “a major move by China to answer the call of the Global South”. Its founding membership – including Cuba, Ethiopia, Kenya, Laos, Pakistan, Brazil, Indonesia and Venezuela – speaks for itself, as does the attendance of UN Secretary-General António Guterres at the signing ceremony. Even the South China Morning Post acknowledges that Beijing is positioning itself as the advocate of openness “at a time when access to advanced technologies is increasingly constrained by export controls and corporate blacklists”.
Four initiatives, one vision
WAICO is the institutional expression of the Global AI Governance Initiative Xi proposed in 2023, and it takes its place within the broader framework of China’s four global initiatives: the Global Development Initiative (2021), the Global Security Initiative (2022), the Global Civilisation Initiative (2023) and the Global Governance Initiative (2025) – the operational programme of the community with a shared future for humanity. Indeed, the GGI concept paper explicitly identified the governance vacuum in artificial intelligence, alongside the underrepresentation of the Global South, as among the central failures of the existing international system. WAICO is the answer to that diagnosis: international rules on AI jointly formulated rather than imposed.
The speech maps directly onto the four initiatives: AI as “an important driver for shared prosperity” (development); AI as a contributor to “common security” rather than an arms race (security); AI in the service of “mutual learning between civilisations” (civilisation); and a “consensus-based global governance framework” built on “true multilateralism” and the central role of the United Nations (governance).
Beneath the mapping lies a single question, posed in every domain the initiatives touch: is the world a jungle in which the strong prey on the weak, or a family with a single shared home? Applied to AI, the first logic – the logic of capitalism and imperialism – produces export controls, corporate enclosure and an arms race; the second – the logic of socialism – produces open source, capacity building and cooperation.
What WAICO embodies is a fundamentally democratic vision of technological development: decisions about humanity’s future made not in the boardrooms of Silicon Valley or the war rooms of Washington, but through “extensive consultation and joint contribution for shared benefit” among sovereign equals. It is, equally, a socialist vision: technology as a public good, serving people before profit, with its benefits extended deliberately to those the capitalist world system has always excluded.
The United States offers the world an AI race with one permitted winner. China offers a symphony in which every country has an instrument.
