The great AI fork: China pushes for AI commons at Summer Davos

Editor's note: Lin G. is a CGTN economic commentator. The views expressed in this article are the author's own and do not necessarily reflect those of CGTN.

As the World Economic Forum's 17th Annual Meeting of the New Champions — better known as Summer Davos — gets underway in China's Dalian, a fundamental choice confronts global leaders: one that will determine whether artificial intelligence becomes humanity's great equalizer or its most formidable divider.

The timing could not be more urgent. Artificial intelligence is reshaping industries and societies at exponential velocity. Yet paradoxically, our global technology landscape grows more fractured. Inside the United States alone, rival AI giants compete fiercely, each fortifying proprietary fortresses with closed-source models and exclusive data pipelines. Extend this logic globally, and the nightmare scenario writes itself: a world carved into incompatible technological silos, where developing economies are permanently consigned to digital dependency, forced to rent intelligence from the few gatekeepers who built the walls.

Against this backdrop, the 2026 Summer Davos Forum held in China has embraced "Innovating at Scale" as its defining theme. And the question hovering over every panel, every corridor conversation, every multilateral meeting among leaders and professionals from nearly 100 countries, is this: Can we scale innovation without surrendering equity or excluding the majority?

There is, in fact, a path forward — but it demands we confront the most uncomfortable question that technology leaders prefer to avoid: Who captures the dividends of AI, and to whom should they flow?

Two philosophies, two futures

Let us be clear-eyed about the alternatives. The closed-source trajectory championed by dominant American AI firms is, from a narrow capital-return perspective, eminently rational. Proprietary models generate scarcity. Scarcity commands premium pricing. Premium pricing maximizes shareholder value. The logic is impeccable—if one accepts that technology exists primarily to enrich its creators.

China has largely chosen a different route. Not out of altruistic naivety, but from a strategic conviction that artificial intelligence, given its transformative potential, functions less like conventional software and more like a public utility — a foundational layer upon which entire economies and societies will be built.

The Chinese approach centers on open-source ecosystems: foundational models, core architectures, and development frameworks made freely available to any nation, institution, or individual willing to contribute. It enables an agricultural researcher in Africa to fine-tune a model for local crop conditions, a community health worker in Southeast Asia to adapt diagnostic tools for rural clinics, and an educator in South America to build language-appropriate learning assistants — all without seeking permission from any corporate headquarters.

Critics will object that this approach naively ignores competitive realities, that it surrenders first-mover advantage to adversaries who will happily exploit openness while building their own defenses. But this objection fundamentally misunderstands the nature of the goods in question. Artificial intelligence is not a zero-sum game where my gain requires your loss. Its value compounds with participation; a model trained on diverse global data, refined by distributed expertise, and deployed across varied contexts becomes more robust, more adaptable, and ultimately more valuable for everyone.

This is what scalable innovation truly means: not scaling a single company's market share, but scaling participation itself.

Consider the alternative scenario. If every major economy embraced closed, proprietary development, artificial intelligence would rapidly devolve into a patchwork of incompatible kingdoms. Developing nations, lacking both capital and the installed engineering base to compete, would find themselves locked into vendor relationships that extract perpetual rents while offering limited sovereignty over their own digital futures.

Capability without limits, values that bind

Here, a distinction must be drawn with clarity: We should never allow human capability to constrain the upper bounds of AI performance. But human values must constitute the boundaries within which AI operates.

We can pursue maximum technical achievement while simultaneously insisting that the fruits of that achievement be distributed with justice. This is precisely where the divergence between closed-source capitalism and open-source commons becomes most evident.

Yet even the most generous open-source architecture, left to its own devices, will not automatically produce equitable outcomes. The governance superstructure must evolve in parallel.

Who decides what constitutes acceptable AI use? How do we prevent data colonialism—the extraction of valuable local knowledge without fair return? How do we ensure that the Global South participates not merely as consumers of AI products but as co-creators of AI infrastructure? These are not questions that code alone can answer. They require multilateral dialogue, shared rule-making, and global mechanisms that give developing economies genuine agency.

The Summer Davos Forum in Dalian, convening representatives from over 90 countries and regions, represents precisely the kind of platform where such conversations can advance. There is no shortage of technical capability among Chinese AI enterprises and institutions; what distinguishes their posture is an openness to partnership that transcends mere commercial calculation. The message from Dalian is unmistakable: Come, collaborate, co-build. This commons has room for all.

The moment of choice

We stand at a fork in the technological road. One path leads toward fragmentation, rent-seeking, and the replication of global inequality in digital form. The other path—scalable, open, governed by human values while pushing technical boundaries—leads toward a future where AI serves as humanity's common inheritance.

China's open-source commitment is not a concession; it is an investment in a future where innovation scales not by erecting walls, but by tearing them down.

The code is open. The invitation stands.