China Reverses Meta’s AI Startup Acquisition Amidst Growing Scrutiny
China’s National Development and Reform Commission (NDRC) has ordered Meta to dissolve its acquisition of the Chinese AI startup Manus. This regulatory intervention signals an increasingly stringent oversight of foreign investments within China’s burgeoning artificial intelligence sector. The commission explicitly stated its intent to “prohibit foreign investment in Manus in accordance with laws and regulations, and requires the parties involved to withdraw the acquisition transaction.” The deal, initially announced in late December 2025 with an estimated valuation of $2 billion, faced immediate investigation by China’s commerce ministry the following month.
Key Takeaways
- Meta’s $2 billion acquisition of AI startup Manus has been ordered to be unwound by China’s NDRC.
- The intervention highlights increased regulatory scrutiny on foreign investments in China’s AI sector.
- Manus co-founders faced travel restrictions during the regulatory review process.
- The startup had already begun scaling down its China operations prior to the NDRC’s final decision.
The regulatory pressure intensified by March 2026, leading to Manus co-founders Xiao Hong and Ji Yichao being summoned for meetings in Beijing and subsequently barred from departing the country, as reported by Reuters. This action preceded the NDRC’s official order, underscoring the gravity of the situation for both Meta and the startup. Manus, which develops AI agents described as “truly autonomous” for independent task planning and execution, had already initiated a wind-down of its China-based operations, including office closures and employee layoffs, in July 2025. Despite relocating its headquarters to Singapore around mid-2025, Chinese regulators maintained jurisdiction over the acquisition.
Manus had experienced rapid growth, attracting significant investor interest. The startup secured $75 million in funding led by Benchmark in May 2025 and achieved $100 million in annual recurring revenue by December 2025, just eight months post-launch. The NDRC’s direct involvement, as the body responsible for economic planning and AI policy, emphasizes the strategic importance Beijing attributes to advancements in artificial intelligence. Even as approximately 100 Manus employees transitioned to Meta’s Singapore offices in March 2026, the regulatory veto on the acquisition remained firm.
This development occurs as Meta, the parent company of Facebook and Instagram, intensifies its AI ambitions to compete with industry leaders like OpenAI, Anthropic, and Google. The company has recently announced significant internal restructuring, including substantial job cuts and plans to invest heavily in AI chips. The forced divestiture of Manus represents a significant hurdle in Meta’s strategy to acquire cutting-edge AI talent and technology from China.
Long-Term Technological Impact Analysis
The NDRC’s decision to reverse Meta’s acquisition of Manus has profound implications for the future trajectory of AI development and cross-border technological collaboration. For China, this move solidifies its stance on retaining control over strategically important AI assets, potentially fostering a more independent domestic AI ecosystem. This could lead to accelerated development within China, shielded from direct foreign acquisition and aligned with national strategic objectives. The scrutiny on Manus may also serve as a precedent, increasing the complexity and risk for foreign companies seeking to acquire Chinese AI startups, potentially rerouting investment flows towards domestic partnerships or different geographical regions.
From a global perspective, this event underscores the growing geopolitical tensions surrounding AI dominance. It highlights the challenges companies face in integrating decentralized or globally distributed AI talent and technology, particularly when operating in or acquiring from markets with strong nationalistic technological policies. For Web3 and blockchain development, which increasingly relies on AI for smart contract analysis, decentralized AI networks, and user experience enhancements, this regulatory environment could influence the accessibility of advanced AI models and talent. Innovations in Layer 2 scaling solutions and decentralized autonomous organizations (DAOs) might need to adapt to a landscape where AI capabilities are increasingly siloed by national interests, potentially driving demand for more open, permissionless AI integration frameworks that can operate independently of traditional corporate or national boundaries.
Based on materials from : decrypt.co