China Orders Meta to Unwind $2B AI Startup Acquisition in 2026: National Security Breakthrough

China Orders Meta to Unwind $2B AI Startup Acquisition in 2026

In a landmark move that sent shockwaves through global tech markets, China’s Ministry of Commerce ordered Meta to unwind its $2 billion acquisition of AI startup Manus on April 4, 2026. The decision, the first of its kind targeting a foreign tech giant’s AI acquisition, marks a defining moment in the battle for AI sovereignty — where national security now overrides commercial deals, even those backed by billions.

How Manus AI Triggered National Security Concerns

Manus, originally founded in Beijing by Chinese nationals, developed an autonomous general AI agent that achieved an unprecedented 86% score on the GAIA benchmark — the highest ever recorded. Though it relocated its headquarters to Singapore in 2025 to navigate U.S.-China tensions, Chinese regulators determined that its core research, engineering team, and funding remained rooted in China. The Ministry of Commerce concluded the move was a deliberate attempt to circumvent domestic oversight, classifying the deal as a "conspiracy" to export frontier AI technology without a license.

Exit Bans and the New Era of Tech Sovereignty

In an unprecedented escalation, Chinese authorities imposed exit bans on Manus’s three co-founders, including founder Xiao Hong, preventing them from leaving the country. This tactic, previously used in high-profile IP theft cases, signals a new regulatory frontier: personal control over technological assets. Chinese officials cited national security concerns, noting that Manus received partial funding from domestic grants and that its algorithms were developed by engineers based in Beijing.

Global Precedents: China’s CFIUS-Style AI Review

China’s action mirrors U.S. CFIUS (Committee on Foreign Investment in the United States) reviews but with sharper teeth. While the U.S. blocks acquisitions based on risk assessments, China now proactively claims jurisdiction over any AI firm with Chinese origins — regardless of incorporation. Regulators are drafting new legislation requiring pre-approval for all foreign acquisitions involving Chinese-founded AI startups, effectively creating a "technology diaspora" screening regime. Analysts compare it to the U.S. restrictions on semiconductor exports to China, but reversed.

Impact on Foreign AI Investment and Global Tech

The ruling has already chilled foreign investment in Chinese AI ventures. Venture capital firms are now wary of backing startups with Chinese roots, fearing sudden regulatory reversal. Meta, which planned to integrate Manus’s autonomous AI agents into WhatsApp, Instagram, and its flagship Meta AI assistant, must now rebuild capabilities internally — a costly and time-intensive process. Meanwhile, Chinese regulators are accelerating domestic AI development, redirecting funds to homegrown alternatives like Alibaba’s Tongyi and Baidu’s ERNIE Bot.

What This Means for the Future of AI

This case establishes a new global norm: AI innovation is no longer purely commercial. Sovereignty over AI now transcends borders, and nations will use legal, personal, and economic tools to retain control. Startups may soon face a binary choice: remain under national oversight for local scale, or risk losing access to foundational talent and IP if they go global. The era of borderless AI collaboration may be over — replaced by a fragmented landscape defined by national interests.