China Blocks Meta’s $2 Billion Manus Acquisition, Bars Founders From Leaving Country
Beijing's veto of Meta's largest China AI deal marks a strategic shift from semiconductor restrictions to active gatekeeping of frontier AI talent and intellectual property.
China’s National Development and Reform Commission has blocked Meta’s $2 billion acquisition of Manus, a Beijing-founded AI startup, citing technology leakage risks to the United States—the first time Beijing has vetoed a major foreign acquisition of a Chinese AI company on national security grounds.
The NDRC barred Manus co-founders Xiao Hong and Ji Yichao from leaving China in March 2026, according to Reuters, following a multi-agency investigation that branded the deal a “conspiratorial” attempt to hollow out China’s technology base. The decision comes despite Manus having relocated its headquarters from Beijing to Singapore in June 2025 and laying off its Beijing staff the following month.
Meta announced the acquisition in December 2025, representing its largest China AI investment since regulatory tightening began. Manus had raised $75 million at a $500 million valuation in April 2025, led by Benchmark with participation from Tencent, HongShan Capital, and ZhenFund, per Today in AI. The company claimed over $100 million in annualized revenue within eight months of launching its flagship agentic AI product in March 2025.
From Semiconductor Walls to Software Gatekeeping
The Manus block represents a fundamental escalation in Tech Decoupling. Where previous restrictions focused on semiconductor equipment and chip exports, Beijing is now actively preventing the transfer of frontier AI capabilities—including trained models, engineering talent, and architectural knowledge—to American companies.
“Advanced AI agents, models and related IP are strategic assets that China considers strategically sensitive.”
— Nick Patience, AI Lead, The Futurum Group
The NDRC’s intervention follows a broader policy shift. In late April 2026, the commission instructed major Chinese AI firms—including Moonshot AI, StepFun, and ByteDance—to reject U.S. capital without explicit government approval, according to Bloomberg. This reverses years of policy encouraging foreign investment in China’s tech sector.
The timing aligns with China’s accelerating AI capabilities. Stanford’s 2026 AI Index shows the U.S.-China performance gap has shrunk from double digits in 2023 to near parity, with models like DeepSeek-R1 matching leading American systems on key benchmarks, according to IBTimes. As China’s domestic AI capabilities approach the frontier, Beijing appears increasingly willing to forgo foreign capital in exchange for retaining indigenous innovation.
Washington’s Response: Distillation Accusations and Potential Retaliation
The White House has escalated rhetoric around Chinese AI development. On 23 April 2026, Michael Kratsios, Director of the Office of Science and Technology Policy, accused China of conducting “industrial-scale distillation campaigns” to replicate U.S. AI models, stating, “There is nothing innovative about systematically extracting and copying the innovations of American industry,” according to CNBC.
Model distillation refers to training smaller, more efficient AI systems by using the outputs of larger models as training data—a technique that can replicate capabilities without access to original training datasets or architectures. U.S. officials allege Chinese firms are using this method to reverse-engineer American AI systems.
The Treasury Department had already begun scrutinizing Chinese AI investments. Benchmark’s April 2025 funding of Manus underwent review under 2023 outbound investment restrictions, though no public resolution has been announced. CFIUS is now expected to intensify reciprocal restrictions on Chinese tech investments in the United States.
Meta has maintained the transaction “complied fully with applicable law” and anticipates “an appropriate resolution to the inquiry,” according to Reuters. However, the exit ban on Manus founders suggests Beijing views the case as closed rather than under negotiation.
China’s Domestic Champion Strategy Accelerates
The Manus decision fits within China’s broader shift toward AI self-sufficiency. Domestic AI chips now comprise 41% of China’s AI semiconductor market, with roughly half supplied by Huawei, according to the Brookings Institution. This indigenization extends beyond hardware to encompass the entire AI stack—from foundation models to agentic systems like those developed by Manus.
Beijing’s approach differs fundamentally from U.S. Export Controls, which restrict outbound technology flows. China is now restricting inbound capital and outbound talent—effectively ring-fencing its AI sector from foreign acquisition regardless of where companies are legally domiciled. Manus’s Singapore headquarters offered no protection once Beijing decided the underlying technology and talent were strategic assets.
What to Watch
CFIUS is likely to announce reciprocal restrictions on Chinese AI investments in U.S. companies within 60 days, particularly targeting early-stage venture capital. The Treasury Department’s unresolved review of Benchmark’s Manus investment may produce the first forced divestiture under outbound investment rules.
Monitor whether other U.S. tech companies with pending Chinese AI acquisitions—particularly in computer vision and natural language processing—face similar blocks. If the NDRC applies the Manus precedent broadly, the window for cross-border AI M&A may have closed entirely.
The exit ban on Manus founders sets a concerning precedent for dual nationals and returnees in China’s tech sector. Entrepreneurs who previously moved freely between Silicon Valley and Beijing now face the risk of being trapped if their companies are deemed strategically sensitive. This will likely accelerate the exodus of Chinese AI talent to Singapore, the UAE, and other neutral jurisdictions—or discourage international collaboration altogether.