USA and USSR did after World War II. Deep mistrust prevents the alliance that Friedman calls for. But AI Middle Way will reduce this mistrust and produce new incentives on both sides as they look beyond easy-to-serve markets in the Global North to forge new markets for the Global South.​
Q: Haven't the nations of the Global South favored one side or another in a new cold war between the US and China?
No. The biggest influential nations of the Global South with billions of consumers—India, Indonesia, Brazil, South Africa, Mexico, and most others in Latin America—have each remained neutral in the AI struggle between the USA and China. Their neutrality gives them leverage to cause market forces to align both superpowers into the shaping of AI ecosystems in the South. ​​​​​​​​
​
Q: What drives the AI Middle Way?
​
Philosophy. It has emerged from the philosophy department of Thailand's leading university, Chulalongkorn. The term "middle way" is rooted in Buddhist philosophy but is equally applicable to Christianity and Islam.. According to the famed economist Kenneth Boulding, "middle way" is not just the midway point between two extremes. A middle way is an emergent property that draws from the best of both sides, creating an altogether different property. ​
​
​
Q: Which nations are involved?
​
The notion began in interaction with AI policy makers in Thailand, Indonesia, Peru and Mexico but was later extended to Brazil and South. Beginning with an online event on March 19, 2026, policy makers in these six proof-of-concept nations will sign a Declaration on AI Middle Way. To see the prediction of how the future of these nations will change if they adopt AI Middle Way – or remain in the grip of data colonialism, go here. These nations represent approximately 600 million people and embody the values of sovereignty and technological self-determination.
​​
​ Q: How does the AI Middle Way engage with the Global South?​
The challenge is to consider best practices from these proof-of-concept nations as well as funding, innovation and technology-transfer from the USA and/or China. For one example from Indonesia, for how this will work, go here.
​
Q: Will the United Nations be driving this?
No. The UN will have an important role in its implementation, but its record on climate (which is now 10% hotter despite the UN embracing the issue with unachieved aspirations) shows that its civil service system is too cumbersome, too slow, too underfunded, and lacks the expertise to move quickly on this urgent issue.
​​
Q: If not the UN, what other intergovernmental agency could take the lead on AI Middle Way?
It is the Group of 20 (G20). The G20's social impact has been strengthened in 2025 by the addition of the 75-nation African Union. Furthermore, by historic happenstance, the four big developing nations of the Global South have recently hosted the G20—Indonesia (2022), India (2023), Brazil (2024), and South Africa (2025)—thereby helping to move its focus away from the Global North.
​
Q: But isn't the G20 just a "talk shop" without its own secretariat? Doesn't its framework have to start from scratch each year as a new host takes charge?
That has been a valid criticism in the past. But beginning when Indonesia took charge, the technology ministers of the G20 created a coalition called the Digital Economy Working Group (DEWG) of the G20. It pledged year-to-year continuity in its deliberations. By 2025 in Johannesburg, DEWG adopted a framework driven by AI for the Global South. This framework is, in essence, the AI Middle Way.
Q: Where does the funding come from for AI Middle Way?
​
It has an eight-tiered macroeconomic model that you can see here is designed to produce the 1.7 trillion dollars billions needed to overcome infrastructural constraints—compute, data centers, electrification, and consumer-market development needed for AI Middle Way to succeed.
Q: Where does the first money come from?​
​​
It comes from US-based philanthropic foundations, which since 1999 have formed coalitions to "close the digital divide" in developing nations. Updated in the new AI era, these foundations—especially Ford, McGovern, and Omidyar—have led the way in funding AI for the Global South. They have recently funded DEWG, enhancing its capacity and demonstrating commitment to this vision.
​
Q: How do China and USA and their key frontier models of AI such as DeepSeek and Chat GPT get drawn into AI Middle Way?
​
Populous countries such as India, Indonesia, and Brazil have the leverage to use USA against China to "up the ante" for the involvement of both USA and China and their AI companies. Their market size and strategic importance create compelling incentives for both superpowers to participate
constructively.
​
Q: How do open source models such as Facebook's Llama and DeepSeek get activated by AI Middle Way without unleashing chaos?
This is one of the selling points for for AI Middle Way. The answer lies in embedding open source models within governance frameworks that emphasize responsible deployment, capacity building, and alignment with each nation's development priorities. Open source becomes a tool for democratization rather than destabilization.
​
Q: How does this relate to Trump?
Currently Trump leads the G20 on a unilateral, transactional basis. But by November 2026, the host of the G20 moves to the United Kingdom, which began the AI Safety movement in 2023. The UK will push back against the unilateralism of Trump 2.0 to foster multilateralism and fulfill the plans made in 2025 by DEWG, thereby leading to full activation of AI Middle Way. For example, as a member of G20, Saudi Arabia may provide ample infrastructure funding and help the Middle East achieve MBS's dream of leading a Middle East regional digital economy.
​
Q: Clearly, AI Middle Way is an ethics-driven framework. Does it aim to serve the poorest of the poor and embrace the theme "AI for all"?
​
​No. Rather than put forward an unachievable aspiration of "AI for all," the AI Middle Way's demographic focus is to serve the tier of national markets that lies just under the threshold of middle-class urban markets. It targets the lower middle class of the Global South—these number 2.1 billions with incomes averaging $5,000 to $10,000 (2025 purchasing power parity). Most of them are in the bridge between the urban educated middle class and the rural "informal economy." Most have internet access. They have basic (mostly used) cell phones. They have secondary education and enough incentives to activate wealth in AI Middle Way regulations. ​
This demographic represents both the greatest potential for human dignity advancement and the greatest risk of being left behind in the AI era.
​
Q: AI Middle Way is clearly an ethics-driven framework. Is there a leader who could be the moral force for this framework?
Yes. It may be Pope Leo XIV, who spoke to the Conclave in May 2025 to assert that the lifelong aim of the young 69-year-old Pontiff would be to reach out beyond the Curia (Vatican bureaucracy) to activate AI to bring dignity and upskilling to the world's vulnerable citizens, as the Leos had done for centuries before him. His vision aligns perfectly with the AI Middle Way's commitment to human flourishing and justice and enables the implementation of his recent AI Encyclical.
​
The Path Forward
The AI Middle Way represents a third path—neither the market-driven model of the USA nor the state-controlled model of China, but an approach that leverages the strengths of both while serving the genuine development needs of the Global South.
The window for establishing this framework is narrow. As AI infrastructure lock-in becomes irreversible by late 2027, the year 2026 represents the final critical opportunity to establish governance mechanisms that ensure AI serves human flourishing rather than concentrating power and wealth.
The framework is emergent. The coalition is forming. The moral imperative is clear.
The AI Middle Way is not inevitable—but it is possible. And it may be the most important policy innovation of our time.
For more information, visit www.middleway.ai
The AI Middle Way Coalition includes Chulalongkorn University (Thailand), government partners in Indonesia, Brazil, Mexico, and Peru, and philanthropic support from leading US foundations committed to global development.
Matt Sheehan, Carnegie Endowment for International Peace: “The West should not just dismiss China’s AI regulations for the making of a surveillance state. Its AI regulations arre nuanced and evolving. For its own benefit, U.S. should "actually understand what it is competing against.”
NYT’s Thomas Friedman: “AI is the danger that will unite both USA and China, despite their being bitter enemies. They will cooperate at a depth they never attempted before even as they compete. Through “co-opatition,” they will prevent the worst AI outcomes: deepfake warfare, autonomous systems going rogue, or runaway misinformation machines."
Luis Ignacio Lula da Silva, President of Brazil “Brazil will not choose sides between USA and China. Our policy is Active Engagement towards both sides”