Animating the Belt and Road Through Digital Humans

The Belt and Road Initiative is China's most ambitious campaign of global connectivity, extending economic, political, and cultural influence across Asia, Africa, Europe, and beyond since Xi Jinping announced it in 2013. In recent years it has grown a digital limb, the Digital Silk Road, which carries the same ambitions into the domains of networks, platforms, and data. It is tempting to fold China's parallel rise in digital humans—AI-driven virtual beings that speak, gesture, and hold a face—into this story as a new instrument of soft power, a synthetic diplomat capable of shaping foreign preferences through appeal rather than coercion. The temptation should be resisted. Whatever their eventual geopolitical weight, digital humans intersect with the Belt and Road far more through commerce than through statecraft, and the connection is one that analysts have largely inferred rather than found. No Chinese policy document connects the technology to the initiative at all.

The Digital Silk Road, formally proposed at the first Belt and Road Forum in May 2017, is built almost entirely from hard infrastructure: 5G networks, submarine cables, data centers, cloud platforms, mobile payment systems, and the e-commerce rails that move goods across borders. By 2022 China had signed dedicated memoranda with seventeen countries and established Silk Road e-commerce mechanisms with twenty-three, and an estimated seventy-nine billion dollars has flowed into such projects through more than 1,300 overseas deployments. The state's planning documents track this physical emphasis faithfully. The 15th Five-Year Plan, approved in March 2026, devotes a full chapter to co-building the Belt and Road and expands cooperation in green development, artificial intelligence, and the digital economy—yet it never mentions digital humans, and neither did its predecessor. The initiative's documented deliverables are towers, cables, and server halls. Virtual beings appear in none of them.

This absence is not for want of government interest in the technology itself. China's support for digital humans is real and considerable, but it points inward, toward domestic industry. Beijing issued the world's first dedicated digital-human industrial policy in August 2022, an action plan aiming to build a sector worth more than fifty billion yuan by 2025, and the first national document to name digital humans repeatedly, the metaverse industry three-year action plan of September 2023, called for cultivating benchmark products, tool components, and well-known virtual figures. Every application it imagined—guides, customer-service agents, shopping assistants—was domestic. Standards work led by the China Academy of Information and Communications Technology has drawn in Tencent, Baidu, Alibaba Cloud, Huawei, and others, but it too is oriented toward ordering a home market rather than projecting abroad.

The same inward orientation governs regulation, where China has built the most developed rulebook for synthetic humans anywhere in the world. The Deep Synthesis Provisions, effective January 2023, require conspicuous labeling of generated content and consent for editing a person's face or voice; the Generative AI Interim Measures that August became the country's first binding comprehensive regime for the field; content-labeling rules and a national standard followed in 2025; and in April 2026 the Cyberspace Administration published a draft regulation specifically for digital-human services, mandating disclosure, explicit consent for any use of a likeness, and prohibitions on using avatars to bypass identity verification or form virtual intimate relationships with minors. Yet nothing in this architecture reaches across a border. No mechanism obliges a Chinese firm to apply these rules when it deploys a digital human abroad, and China promotes its governance model internationally through diplomatic initiatives and multilateral forums, not by extending its regulations into partner countries.

What does cross borders is the industry. China's digital-human market, valued at more than four hundred million dollars in 2024 and projected toward several billion within a decade, supports well over a thousand active companies, and the strongest of them have begun looking outward for growth. Silicon Intelligence, a Nanjing firm holding roughly a third of the domestic agent segment, has deployed tens of thousands of avatars and filed to become the sector's first dedicated public listing; Baidu cut the cost of a hyper-realistic 3D digital human to under two hundred yuan; SenseTime's avatar system spans the languages of more than a hundred countries. These firms are moving into Southeast Asian and Middle Eastern markets not because a directive tells them to animate the Belt and Road, but because those markets offer the fastest growth and Chinese companies are best positioned to serve them.

That commercial logic produces the densest real intersection between digital humans and Belt and Road geography, and it runs through cross-border e-commerce livestreaming. Alibaba's international arm introduced a registration system for digital-human livestreams in early 2024, and control platforms such as BocaLive now drive synthetic presenters across Shopee, TikTok, Lazada, and other marketplaces in a dozen languages including Malay, Indonesian, Thai, and Vietnamese. The documented results are concrete—thousands of monthly orders for a children's-clothing seller on Shopee, a single livestream in Indonesia clearing billions of rupiah in sales—but the motive is plainly economic: lower labor costs, round-the-clock multilingual operation, and content that scales without end. These streams happen to reach markets that overlap with Belt and Road corridors; they are not steered there by anyone. Tellingly, the trend has already begun to cool as platforms learn to detect and penalize automated broadcasters.

The handful of cases that come closest to statecraft turn out, on inspection, to be showcases and sponsorships rather than coordinated diplomacy. At the 2025 China-ASEAN Expo, iFlytek fielded the event's first digital-human hosts alongside real-time interpretation across more than a hundred languages—the clearest verifiable meeting of a specific product with a named Belt and Road event, and still a commercial partnership. In cultural programming, an ultra-realistic virtual performer inspired by the Dunhuang flying apsaras appeared at the 2023 Silk Road Spring Festival Gala, weaving a piece of intangible heritage into a broadcast that drew billions of media impressions. State broadcasters have built multilingual AI news anchors, and Xinhua runs a Belt and Road information service, so the infrastructure for themed content plainly exists; what is missing is any evidence that it has actually been turned to that purpose.

Education tells a similar cautionary story about reading too much into a label. China's flagship vocational-training brand along the Belt and Road, the Luban Workshop, now operates in dozens of countries, and its classrooms are genuinely advanced—but the technology inside them is virtual-reality simulation and smart-classroom equipment, not AI avatar instructors. The workshops in Kyrgyzstan, Thailand, and Kenya run immersive laboratories and cross-border VR teaching rather than synthetic teachers, and the distinction matters for anyone trying to gauge how digital humans are, and are not, embedded in the initiative's programming.

The stakes are real enough that a substantial critical literature has grown up around Chinese digital exports, warning that Digital Silk Road infrastructure creates dependencies and hands Beijing leverage over governance norms in less-developed regions. That critique deserves a hearing, and so do its rebuttals: analysts note that African and Southeast Asian states exercise considerable agency, deliberately diversifying across competing vendors, and that countries such as Kazakhstan have devised their own digital strategies partly as a hedge against Chinese influence. What no one has produced is a study of digital humans specifically in Belt and Road partner countries. To cast them as instruments of digital colonialism today is to extrapolate from the broader infrastructure debate, not to report a finding.

The honest description, then, is of an emergent commercial phenomenon that overlaps with the Belt and Road's map rather than a deliberate projection of technological soft power. The more productive question is not whether China is deploying digital humans as agents of influence—there is no evidence that it is—but whether their commercial diffusion into these markets will raise the same issues of data dependency, content sovereignty, and governance norms that already shadow the Digital Silk Road. As the technology grows cheaper and more convincing, those questions will sharpen on their own, whether or not any policy document ever instructs a digital human to speak on behalf of the Belt and Road.