The $2 Billion Exit That Never Was
Two founders built something Meta wanted badly enough to pay $2 billion for. Then Beijing stepped in. Here is what that story means for anyone building a digital business right now.
Red Xiao and Peak Ji spent years building Manus AI, an agentic system capable of completing complex tasks on its own. Not just answering questions. Actually doing work. The product was sophisticated enough that Meta offered two billion dollars to acquire it. That number is not a typo.
The deal was announced in December 2025. By March 2026, both founders were reportedly barred from leaving China. By late April, Beijing's National Development and Reform Commission had cancelled the acquisition entirely. The exit they spent years working toward was gone.
The summit for marketers who are tired of guessing.
Search has changed more in the last 18 months than the previous decade. Most teams are still running yesterday's playbook.
The Agentic Marketing Summit changes that. 3x Inc 5000 founder and search expert Manick Bhan breaks down the exact science behind getting found online.
Real Experts. Live. One week only.
This is not a story about geopolitics. At least, not entirely. It is a story about what happens when the rules change faster than your business plan.
For years, the playbook for Chinese tech founders was well understood: build something world-class, attract Western capital, and exit to a US tech giant. Singapore gave you a neutral flag to sail under. The deal would close, the wire would transfer, and you would move on to the next thing.
That playbook is now obsolete.
Beijing decided that Manus AI, despite operating under a Singapore holding company, was still Chinese intellectual property. The founders had Chinese roots. The engineering talent was predominantly Chinese. In the government's view, selling to Meta was equivalent to exporting a national asset without permission. The Singapore incorporation was not a workaround. It was irrelevant.
The NDRC is now functioning as China's version of CFIUS, the US body that reviews foreign acquisitions for national security risks. The difference is directional. CFIUS screens deals coming into the US. The NDRC is now screening deals going out of China, even when the company is not technically based there.
If your founding team or core IP has any connection to China, your exit to a US buyer may no longer be guaranteed, no matter where your company is incorporated.
What the decoupling actually looks like on the ground
The first consequence is already playing out in hardware. Nvidia built a chip specifically for the Chinese market, the H20, designed to stay within US export rules while still generating revenue from Chinese buyers. Beijing has reportedly directed Chinese firms to stop purchasing it altogether and move to domestic alternatives like the Huawei Ascend series.
That is not a small shift. It affects the economics of every data center, every cloud provider, and every AI tool built on top of that infrastructure. If you use any AI-powered software in your workflow, the hardware it runs on is getting more expensive and more fragmented. That cost will eventually show up in your subscription price.
The second consequence is talent flow. The Manus case has created what analysts are calling a "dual-exit trap." Founders with Chinese ties cannot sell to US companies without risking a Beijing block. They cannot raise from US venture capital without triggering scrutiny from both sides. The most innovative AI research in China will stay in China. Two ecosystems are forming, and they will not be compatible with each other.
The opportunity hiding in the friction
Fragmentation creates arbitrage. This is worth sitting with.
When two separate ecosystems develop, there are gaps between them. Tools built for one market do not transfer cleanly to the other. Content strategies that work on US platforms do not map to their Chinese counterparts. The workflows, formats, and integrations differ. Someone has to bridge those gaps, or build natively within each one.
For creators and marketers, this means knowing which market you are actually serving, and going deep into that ecosystem rather than hedging across both. The era of building one product for everyone is narrowing. Specificity is becoming a structural advantage, not just a preference.
For entrepreneurs, it means watching where capital is flowing. Chinese state guidance funds, the government vehicles that direct investment into strategic sectors, are about to pour significant money into domestic AI startups. Those companies will need infrastructure, content, distribution, and professional services. That is a market, even if you are not a founder.
One thing worth doing this week
Audit your tool stack. Not for political reasons. For practical ones. Many of the tools freelancers and creators depend on are built on infrastructure directly affected by this supply chain split. Know what you use, who makes it, and whether there is a viable alternative if the price spikes or the service disappears. That is not pessimism. It is the kind of operational awareness that separates people who last in this business from those who get caught off guard when the ground shifts.
The Manus founders built something worth $2 billion. They just could not take it with them. The world they built for had changed by the time the deal was ready to close. That is the real lesson here, and it applies well beyond AI acquisitions.
