China’s move to block Meta’s acquisition of Manus signals a new phase of AI techno-nationalism, where talent, algorithms, and cross-border deals are becoming national-security concerns.
Beijing is treating its AI ecosystem as a strategic fortress, blocking outflows of talent, IP, and value in a bid to dominate the next technological era. In late April 2026, China’s National Development and Reform Commission (NDRC) delivered a brief directive that sent shockwaves through global tech circles. It ordered Meta Platforms and the agentic AI startup Manus to unwind their roughly $2 billion acquisition. The deal, announced in late 2025, had already seen Manus staff integrate into Meta’s Singapore offices. Yet Beijing stepped in decisively.
Manus, founded by Chinese engineers Xiao Hong and Ji Yichao, had roots in Beijing but relocated key operations to Singapore. This move aimed to tap global capital and ease geopolitical friction. Founders faced travel restrictions during the probe. The NDRC’s intervention wasn’t just about one transaction. It signaled a profound policy shift: AI talent, algorithms, and foundational capabilities linked to China may increasingly be treated as national-security assets, even when companies restructure offshore.
This isn’t isolated protectionism. It marks Beijing’s evolution from welcoming foreign investment in tech to actively sealing off strategic outflows. As the U.S.-China rivalry intensifies over artificial general intelligence (AGI) pathways, agentic systems, and multimodal models, China is weaponizing its regulatory toolkit. No longer content with shielding its domestic market, Beijing now prevents the exodus of what it views as core competitive advantages. This escalation deepens the tech decoupling and forces a fragmented global AI landscape.
The Shifting Sands of Chinese Tech Policy
China’s approach to foreign investment in technology has undergone a dramatic transformation. In the 2010s, the country aggressively courted FDI, red-chip listings, and Western venture capital to fuel its tech rise. Billions flowed into Alibaba, Tencent, and a generation of startups. Yet cracks appeared by the late 2010s, widening sharply after 2021.
That year, China implemented the Measures for Security Review of Foreign Investment. These rules empowered the NDRC and Ministry of Commerce to scrutinize deals affecting national security, covering critical sectors like information technology, key technologies, and data. AI quickly fell under this umbrella, viewed alongside semiconductors as a foundational pillar of future power.
Post-2024 acceleration came amid U.S. export controls on advanced chips, outbound investment restrictions finalized in early 2025 targeting AI and quantum, and growing fears of talent drain. Chinese founders increasingly explored “Singapore-washing”-relocating headquarters or intellectual property offshore to attract U.S. or global buyers while distancing from Beijing’s oversight. Manus exemplified this strategy. Its relocation, initially seemingly tolerated, ultimately backfired spectacularly.
Broader patterns reinforce the trend. Regulators have instructed prominent AI firms-including Moonshot AI, StepFun, and even ByteDance – to reject U.S.-origin capital in funding rounds without explicit approval. Scrutiny of red-chip structures and secondary share sales has intensified. Beijing now asserts jurisdiction over technology, talent, and data ties to China, irrespective of incorporation location.
Why Beijing Is Drawing These Red Lines
Several interlocking motivations drive this strategy. At its core lies national security fused with technological self-reliance. Chinese leaders see AI, particularly agentic systems capable of autonomous planning, tool use, and multi-step reasoning, the new semiconductor race. Losing control over frontier models, training methodologies, or elite researchers could cede decisive advantage in AGI development. Manus specialized in agentic AI, promising transformative productivity tools. Transferring such know-how to a U.S. giant like Meta was unacceptable.
Geopolitically, this represents tit-for-tat realism. Washington’s chip bans, entity lists, and investment curbs aimed to hobble China’s progress. Beijing responds symmetrically, treating outbound flows as risks to “industrial security.” As one analyst noted, “Beijing effectively drew a bright red line that Chinese AI talent and technology are not for sale to American companies, full stop.”
Economically, the calculus preserves value creation domestically. Early-stage Chinese AI firms have long relied on global exits for liquidity and valuation uplift. By blocking such pathways, Beijing keeps upside-and tax revenues, future champions, and data ecosystems-within its borders. It also deters “shedding Chinese identity,” sending a clear warning: relocation offers no shield.
The signaling effect carries deep intent. Founders contemplating offshore pivots now face regulatory summoning, travel bans, or deal unwinding. This fosters a chilling effect, channeling innovation toward state-aligned priorities like domestic self-sufficiency and applications serving national goals.
Ripple Effects Across the Ecosystem
For Chinese startups, the implications are double-edged. Access to premium global capital becomes harder, potentially slowing scaling for capital-intensive AI training and inference. Talent retention might strengthen short-term as offshore dreams dim, yet long-term innovation could suffer from reduced competition and idea exchange. Many firms may double down on domestic focus, aligning closer with government initiatives but risking echo chambers.
Western tech giants like Meta, Google, or Anthropic encounter heightened risks. Deals with China-origin entities demand exhaustive diligence on ownership, data flows, founder passports, and engineering footprints. Even post-acquisition integration faces reversal threats, complicating talent wars and IP strategies. Meta’s experience-staff already embedded, yet forced to unwind-highlights the operational nightmare.
Globally, this accelerates bifurcation. The AI ecosystem fragments along national lines: separate talent pools, divergent standards, duplicated R&D efforts, and incompatible supply chains. Emerging markets like India stand to gain as alternative hubs, attracting founders seeking neutrality. Singapore’s appeal as a bridge dims under dual superpower pressure.
Investors face rising risk premiums on China-linked AI plays. Valuations may compress without reliable Western exit ramps, while compliance costs soar. Markets will price in heightened geopolitical volatility.
A Realist Perspective on Techno-Nationalism
Critics rightly decry this as self-defeating isolationism that could starve China’s ecosystem of the global collaboration fueling its rapid ascent. Talent thrives on openness; walls may breed complacency or brain drain via subtler routes, like emigration of individuals. In a field advancing through cumulative, borderless insights, excessive control risks turning China into an innovation island-powerful but ultimately limited.
Yet China’s concerns hold legitimacy. The U.S. imposes parallel restrictions via export controls, investment bans, and CFIUS reviews. Both powers recognize frontier AI’s dual-use nature: immense civilian promise shadowed by military applications in autonomy, cyber, and decision superiority. In an era of strategic competition, naive globalization in such domains was always unsustainable. Beijing protects what it built with massive state support, talent cultivation, and data advantages.
Nuance matters. Not every deal triggers blocks, and China continues selective openness. Legitimate security fears around technology leakage coexist with broader industrial policy goals. Mirroring U.S. actions doesn’t make Beijing’s approach flawless; it reveals a shared recognition that unchecked flows in strategic tech invite vulnerability.
Navigating the New Era of Fragmented AI
This Manus episode, paired with capital curbs, heralds a new chapter of techno-nationalism. Borders in silicon and software harden into walls. The race for AGI will unfold in parallel universes-Chinese models optimized for domestic scale and control, Western ones leveraging open alliances and compute access.
Winners will adapt fastest: diversifying talent pipelines, mastering multi-jurisdictional compliance, and innovating within constrained ecosystems. For policymakers, the lesson is clear-smart, targeted safeguards outperform blanket isolation. For founders and investors, due diligence now extends beyond financials to geopolitical mapping.
In this fragmented reality, the illusion of a borderless tech future fades. What emerges is a contested, high-stakes competition where strategic autonomy defines success. China’s moves force everyone to choose sides or build resilient bridges across the divide. The AI arms race has entered its territorial phase, and the contours of tomorrow’s technological order are being etched in regulatory ink today.