One Person, a Thousand AI Employees: How China Built a 12-Million Solo Entrepreneur Army

heroImage: "https://images.unsplash.com/photo-1551650975-87deedd944c3?w=1200"

*Photo: A new kind of company is being born — one where the org chart has a single human name and every other role is filled by an AI agent. China's OPC movement has attracted 12 million solo entrepreneurs. Image: Unsplash*

In May 2026, a photographer visiting an office park in Suzhou captured an image that would have been impossible two years ago. A vast co-working floor, bright with natural light, filled with rows of individual desks. Each desk held exactly one person, one monitor, and one laptop. No standup meetings. No Slack channels buzzing with chatter. No manager hovering. Just twelve hundred solo entrepreneurs, each quietly running an entire company.

This is the OPC — the One-Person Company — and it has become the most consequential economic phenomenon in China's AI era. The term describes a business structure where a single founder uses AI agents to perform every function that once required a team: coding, design, marketing, customer service, finance, legal compliance. The human makes decisions. The AI executes.

The numbers are staggering. By mid-2026, more than 12 million individual entrepreneurs in China had entered the OPC track. Over 20 cities have launched dedicated OPC support policies, competing with subsidies, free office space, and computing power vouchers to attract solo founders. The market these one-person companies serve is projected to exceed 800 billion RMB. And Dario Amodei, CEO of Anthropic — an American company with no stake in China's startup scene — has publicly predicted that the world's first one-person unicorn will emerge before the end of this year.

To understand why this is happening in China first, and why it is happening now, you have to look past the technology. The tools are available everywhere. What is different is everything underneath them.

The Scale of a Movement

The OPC did not appear from nowhere. It is the convergence of three trends that have been building for years: the collapse of AI agent costs, the maturation of China's large language model ecosystem, and a policy environment that has decided — explicitly and with money — that the future of work is one human plus many machines.

Start with the technology cost curve. In 2024, running a full suite of AI agents capable of handling code generation, content creation, customer interaction, and data analysis would have cost approximately $5,000 per month. By mid-2026, that same capability cost $40 per month — a 128-fold reduction in just two years. For the price of a nice dinner, a solo founder can now deploy what amounts to a full digital workforce.

Here is the number that matters most. Because while the technology made OPCs *possible*, it was government policy that made them *inevitable*.

The User Stories: Three Faces of the Revolution

Numbers tell you the scale. People tell you the story. Across China, the OPC movement has produced thousands of individual narratives that would have sounded like science fiction eighteen months ago.

The Technical Founder

Xu Zhanwei holds a PhD from Tsinghua University. In July 2025, he registered a one-person company focused on smart wearable devices. He had deep expertise in interaction technology but no team, no industrial connections, and no business background. What he had was a set of AI tools that let him design product prototypes, generate manufacturing specifications, draft investor presentations, and manage his supply chain communications.

By October, he had settled into Shenzhen's Moli Camp — one of the city's first OPC communities. Within six months, he had raised more than 10 million RMB and secured two major commercial orders. His total headcount remained exactly one.

The Career Changer

Wang Yifeng graduated from Fudan University and spent several years at a state-owned enterprise before quitting in early 2026 to register a one-person content company. She had no coding skills and no design experience. What she had was expertise in her field and an ability to use AI tools that compressed her article-writing time from two to three hours per piece to under one hour.

Within two months, her income slightly exceeded her previous salary. Her business has no employees. Her "team" consists of a Chinese LLM for drafting, an image generation model for illustrations, and an automated distribution pipeline that pushes content to six platforms simultaneously.

The Non-Technical Builder

Kelly is a product manager and Tsinghua graduate with zero programming background. In early 2026, she decided to test a hypothesis: could AI let her build a working product entirely on her own? She spent one week and approximately $5 on AI tool subscriptions. The result was a functional clone of Bing Search — front-end, middleware, back-end, all built by prompting AI systems.

Her conclusion, shared widely on Chinese social media, became an unofficial motto for the movement: *"After one person interacts with AI, it is equivalent to having the capabilities of an entire team including front-end engineers, middleware developers, and back-end engineers."*

These three stories share a common thread: none of these founders could have built their businesses alone in 2024. The AI tools existed, but they were assistants, not employees. What changed in 2025 and accelerated through 2026 is the shift from AI as a tool you operate to AI as a workforce you direct.

Behind the Scenes: The Architecture of a One-Person Company

What does the internal structure of an OPC actually look like? The emerging standard, documented across hundreds of Chinese OPC communities, follows a consistent pattern.

The founder sits at the top of this stack as the sole decision-maker. They define the product vision, set quality standards, handle relationships that require human trust, and make the judgment calls that AI still cannot. Everything else is delegated to what the Chinese ecosystem calls "digital employees" — AI agents with defined roles, permissions, and accountability structures.

*The modern OPC workspace: a single human directing a fleet of specialized AI agents, each handling a function that once required a dedicated employee. Image: Unsplash*

This architecture is possible because of a specific technical convergence that happened in China during 2026. The country's domestic LLM ecosystem reached a critical mass: in February 2026, Chinese LLMs collectively processed 4.12 trillion tokens per week, surpassing American models' 2.94 trillion for the first time. Four of the top five global LLMs by usage were Chinese. The API costs for these models dropped to levels that make large-scale agent deployment economically trivial for individual users.

But the technology is only half the story. The other half is what the government built on top of it.

The Government Bet: Cities Compete for Solo Founders

In March 2026, Shenzhen's Longgang District released a policy document that would have been unthinkable in any other country: the "Ten Measures to Support OpenClaw and One-Person Company Development" — nicknamed the "AI Lobster Ten" (AI龙虾十条), a pun on Longgang (龙岗) and the "shrimp-raising" metaphor that China's OpenClaw community uses for training AI agents.

The policy reads like a wish list for solo founders:

Longgang was first, but not alone. Within months, more than 20 Chinese cities had launched OPC support programs, each trying to outbid the others.

The logic is straightforward from a policy perspective. China faces a structural employment challenge: a large, well-educated workforce and a tech sector that has been trimming headcount as AI adoption increases. Rather than fighting this transition, policymakers have chosen to accelerate it — converting displaced workers into entrepreneurs who use AI to multiply their own productivity. The OPC model does not just preserve economic output; it potentially multiplies it by removing the coordination overhead and labor costs that consume most of a traditional startup's early capital.

CPPCC delegates formally raised the OPC concept during the 2026 Two Sessions, elevating it from a local experiment to a national policy direction. By October 2026, OPC communities — physical spaces where solo founders share office infrastructure, attend AI training sessions, and collaborate on projects — have proliferated across the Yangtze River Delta, Pearl River Delta, Beijing-Tianjin-Hebei region, and emerging industrial cities in central and western China.

Global Parallels: Why China, Why Now

The OPC concept is not uniquely Chinese. Pieter Levels, the Dutch indie hacker who has become the global patron saint of solo entrepreneurship, has been running one-person companies for over a decade. His portfolio — Nomad List, RemoteOK, and others — generates more than $5 million in annual revenue with zero employees. His approach, which he calls "Build in Public," has inspired a global community of indie hackers and solopreneurs.

But the Chinese version of the OPC movement differs from its Western counterpart in three fundamental ways.

The most striking difference is speed. The Western indie hacker community took fifteen years to reach its current size. China's OPC movement grew from zero to 12 million participants in roughly eighteen months. Part of that is the technology maturation curve — Chinese founders had access to more capable AI agents from day one. Part of it is the policy push. And part of it is cultural: the concept of the "super individual" (超级个体) — a person whose capabilities are amplified to match an entire team — resonated immediately with a generation comfortable delegating tasks to algorithms.

Tom van Dillen, managing partner at Greenkern, told CNBC what many Western observers are thinking: *"China is turning an open-source tool into national productivity infrastructure at a speed no other country is matching."*

The contrast with Hong Kong is instructive. According to the Hong Kong Companies Registry, more than 75% of new companies incorporated in 2024-25 were one-member companies. Yet Hong Kong has no equivalent of Shenzhen's Lobster Ten Measures — no computing vouchers, no OPC communities, no dedicated government support. The companies exist, but the ecosystem does not. The result: Hong Kong's OPCs remain lifestyle businesses, while Shenzhen's are building toward unicorns.

The Risks Nobody Talks About

For all its energy, the OPC movement faces real challenges that its boosters prefer not to discuss.

AI hallucination in critical decisions. When an AI agent handles your finances, reviews your contracts, or manages your customer communications, errors compound. A hallucinated invoice number or a misunderstood legal clause can cascade into serious problems. Experienced OPC founders report that they still manually review anything involving money, legal obligations, or customer commitments — which undermines the "fully autonomous" narrative.

Market saturation. With 12 million participants, differentiation becomes increasingly difficult. Generic AI-powered services — content creation, basic app development, chatbot deployment — are commoditized almost as fast as new entrants can launch them. The founders who succeed are those with genuine domain expertise that AI cannot replicate, not those who simply know how to prompt.

Regulatory uncertainty. China's government has been simultaneously supportive and cautious. The Ministry of Industry and Information Technology is developing national standards for AI agents covering permission management, execution transparency, and behavioral risk controls. OPCs built on OpenClaw — which has faced security scrutiny, with reports of tens of thousands of exposed instances — may find their technology stack subject to new compliance requirements.

The loneliness problem. Multiple OPC reports have flagged a paradox: the same autonomy that attracts people to solo entrepreneurship also isolates them. Without colleagues, founders report higher rates of burnout and decision fatigue. The OPC communities that cities are building are partly a response — physical spaces where solo founders can find the social infrastructure that a traditional office provides.

Sustainability questions. The subsidy-driven model means many OPCs are economically dependent on government support. If policy priorities shift, the underlying business models may not be viable without continued public funding. Whether the movement can produce self-sustaining businesses at scale remains an open question.

What Comes Next

The trajectory is clear, even if the destination is not. Current projections suggest China will have more than 20 million OPC entrepreneurs by the end of 2026, and that over 50% of new startups will adopt the solo-founder model by 2028. The first OPC unicorn — a one-person company valued at over $1 billion — is widely expected to emerge within the next twelve months.

Several developments will determine whether the movement delivers on its promise or collapses under its own weight.

The maturation of agent-to-agent protocols will be critical. Today, most OPCs use AI agents that operate in isolation — one for coding, one for marketing, one for customer service. The next generation of agent infrastructure will allow these digital employees to communicate with each other, negotiate, and coordinate complex multi-step projects with minimal human oversight. When that happens, the capacity of a single founder expands from "one-person company" to "one-person conglomerate."

The internationalization of Chinese OPCs is another frontier. Several cities, including Shenzhen Longgang, have already included export support in their OPC policies — one-stop services for cross-border market expansion, logistics, compliance, and export credit insurance. Chinese solo founders building AI-powered products for global markets, subsidized by local governments, could become a significant force in international digital trade.

And the employment question will not go away. If OPCs genuinely allow one person to do the work of fifty, the aggregate impact on China's labor market is profound. The optimistic view is that OPCs will absorb displaced workers and create new forms of economic participation. The pessimistic view is that they represent a transitional phase — a way to manage the decline of traditional employment without addressing root causes. The truth likely lies in between.

The photograph from Suzhou captures a moment in this transformation: twelve hundred people, each alone at their desk, each running a company. It looks like solitude. It is actually a crowd.

*OPC communities across China provide solo founders with shared office infrastructure, AI training programs, and the social connections that independent work often lacks. Image: Unsplash*

These projections carry significant uncertainty. But the direction — toward smaller teams, AI-augmented individuals, and new definitions of what constitutes a business — is not in doubt.

What is certain is that the definition of a "company" is being rewritten in real time, and China — with its 12 million solo founders, its city-level subsidy competitions, and its willingness to treat AI agents as economic infrastructure — is where the rewriting is happening fastest.

The photograph from Suzhou captures a moment in this transformation: twelve hundred people, each alone at their desk, each running a company. It looks like solitude. It is actually a crowd.

Social Voices

@数字游民小陈 (Digital Nomad Xiao Chen) — Zhihu

"我在深圳龙岗的OPC社区办公，隔壁桌做跨境电商，对面做AI配音工具，就我一个人做SaaS。三个人，三家公司，加起来不到10平米。去年我年入80万，成本不到5万。这在以前根本不可能。"

>

"I work at an OPC community in Shenzhen Longgang. The desk next to me does cross-border e-commerce, across from me someone builds AI dubbing tools, and I'm the only one doing SaaS. Three people, three companies, sharing less than 10 square meters. Last year I made 800K RMB with costs under 50K. This would have been impossible before."

@Tech Observer Lisa — Twitter/X

"Everyone in Silicon Valley is debating whether AI will replace jobs. In Shenzhen, they've moved past debating — they're just building companies where AI does the work and one person keeps the equity. The speed difference is not technological. It's philosophical."

@独立开发者老王 (Indie Dev Lao Wang) — Weibo

"OPC不是让你一个人硬扛，而是让你一个人指挥一个AI军团。关键不在于你有多强，而在于你会不会'养龙虾'。会用AI的人，一个顶五十个。"

>

"OPC isn't about one person struggling alone — it's about one person commanding an AI legion. The key isn't how capable you are; it's whether you know how to 'raise lobsters.' Someone who can use AI effectively replaces fifty people."

@Sarah Chen, VC Analyst — Xiaohongshu

"作为投资人，我现在看到BP写着'团队：1人'的时候不会直接pass了。反而会多看两眼。因为如果一个聪明人加上AI工具能把MVP做出来，说明这个人的执行力和判断力都是顶级的。问题是，这种公司怎么估值？传统估值模型完全失效。"

>

"As an investor, when I see a business plan that says 'Team: 1 person,' I no longer pass immediately. I actually look closer. If a smart person plus AI tools can build an MVP, it shows top-tier execution and judgment. The problem is: how do you value a company like this? Traditional valuation models completely break down."

@Skeptical Panda — GitHub Discussion

"I've been running a one-person SaaS for 8 months in Chengdu. Revenue is fine, but the mental toll is real. No one to bounce ideas off. No one to cover when you're sick. AI agents don't give you feedback on strategy — they execute whatever you tell them, even when you're wrong. OPC is powerful but it's not free."

@商业模式观察 (Business Model Observer) — Douban

"政府补贴OPC听起来很美好，但想想之后呢？如果补贴停了，这些一人公司能活多久？很多OPC本质上还是'用AI包装的自由职业'，不是真正的公司。等风口过了，能活下来的可能不到10%。"

>

"Government-subsidized OPCs sound great, but think about what happens after. If subsidies stop, how long can these one-person companies survive? Many OPCs are essentially 'AI-wrapped freelancing,' not real companies. When the hype passes, maybe less than 10% will survive."

*Related reading: China AI Goes Global: The Open-Source Cloud Empire Nobody Saw Coming · China's 700 Million AI Users: The Mainstream Tipping Point · Space Bunny: The Anonymous AI Model Shaking China's Tech Industry · China's AI Creator Economy: The Digital Human Gold Rush*