China's NDRC Halts Meta's $2 Billion Acquisition of AI Firm Manus

China's economic planning body, the NDRC, has effectively halted Meta's $2 billion acquisition of AI firm Manus. Despite the company reincorporating in Singapore and moving its operations, Beijing intervened, barring founders from leaving and ordering the deal unwound. This move signals a significant shift in China's oversight of tech deals, impacting future foreign investments.

Three engineers in Wuhan had a clean plan. Build a world-class AI agent, take foreign money, move the company to Singapore, and sell to the highest bidder. By December, that bidder was Meta — $2 billion, done deal, engineers already settling into Singapore offices. It looked, for a moment, like the kind of breakout story Chinese tech founders dream about. Then Beijing made a phone call. Then it barred the founders from leaving the country. Then it ordered the whole thing unwound. The agency behind that decision isn't a spy bureau or a military commission. It's the National Development and Reform Commission—a 70-year-old economic planning body that used to approve dam projects and set coal targets. It just became the most consequential regulator in global tech.

The NDRC's original job was unglamorous: manage China's long-term economic blueprints, coordinate state investment, keep the lights on. For decades it was powerful in the way that any large bureaucracy is powerful—slow, procedural, and mostly invisible to anyone outside China's planning apparatus. That changed under He Lifeng, Xi Jinping's chief trade negotiator and a former NDRC head who has quietly expanded the agency's reach into territory it was never designed to occupy. The Financial Times reported that the NDRC has shaped decisions on Nvidia chip purchases, intervened in the Panama Canal ports dispute, and now overseen the unravelling of a completed Silicon Valley acquisition. It is operating a foreign investment security review framework that was approved in 2021 but barely used—until now. "The NDRC is emerging as the leading agency of China's version of CFIUS," NYU law professor Winston Ma told the FT. That analogy is useful. America's foreign investment committee has spent years blocking or unwinding deals on national security grounds. China now has a functional equivalent—and the Manus case is its loudest public debut.

Manus followed what had become a well-worn route. After Benchmark led a $75 million funding round in early 2025, the company reincorporated in Singapore, moved its leadership team, shut its Beijing offices, and laid off its China-based staff. The idea was straightforward: shed the Chinese label, become a global company, and make yourself acquirable by Western buyers without the regulatory headaches that come with Chinese ownership. Meta bought it. The money moved. Benchmark distributed returns to its investors. The two teams, by Meta's own description, were "deeply integrated." None of it held up under NDRC scrutiny. Chinese state media accused Manus of "going offshore through washing"—a phrase that has since acquired regulatory weight. The NDRC's position, as the FT reported, was that Manus was built in China, by Chinese engineers, on Chinese soil. Its early R&D happened in Wuhan. Its technical foundation was Chinese. A Singapore address on a company registration didn't change any of that. The co-founders, Xiao Hong and Ji Yichao, were summoned to Beijing in March for talks with regulators. They were later told they could not leave the country.

Cross-border deals between Chinese startups and foreign investors have been collapsing for years—down 73 percent in deal count from the 2021 peak, with total value falling from $54 billion to $7.8 billion, according to PitchBook data cited by the New York Times. The Manus ruling doesn't reverse that trend. It accelerates it. What Beijing established here goes well beyond one acquisition. The NDRC has effectively published a new checklist for any future deal involving a Chinese-origin company: where the early R&D happened, where the engineers lived, where the data was stored, how the offshore restructuring was conducted. Incorporation papers are now largely beside the point. For Chinese founders who built something real, took government support to get started, and hoped to eventually sell to a Western buyer—that exit just became significantly harder to reach, and significantly more dangerous to attempt.