Quick Buzz Feed
China's recent order for Meta to unwind its $2 billion acquisition of AI startup Manus, based on the technology's origin rather than the company's domicile, signals a new, aggressive regulatory stance. This client alert examines Beijing's 'technological nationality' doctrine, its codification through new outbound investment regulations, and the dual regulatory exposure faced by U.S. acquirers of Chinese-origin AI, providing critical advice for in-house counsel on navigating future cross-border tech deals.
Manus, an AI startup founded in Wuhan, China, developed an AI agent framework and later relocated its headquarters and core technical staff to Singapore in a move termed 'Singapore washing.' Meta acquired Manus for over $2 billion without seeking Chinese regulatory approval. Shortly after the deal closed in December 2025, China's Ministry of Commerce launched an investigation, leading to the National Development and Reform Commission (NDRC) issuing a formal prohibition order in April 2026, forcing Meta to dismantle the acquisition. This marked the first public use of China's Foreign Investment Security Review (FISR) mechanism to reverse a consummated AI transaction.
The NDRC's enforcement was distinctive in its analytical framework, referring to the acquisition of the Manus 'project' rather than a specific legal entity. This signaled a focus beyond corporate formalities to the substance of the transferred technology, specifically its origin: where it was developed, where the engineering team gained expertise, and how intellectual property was transferred from the original Chinese entity. Notably, Beijing did not assert current operations in China, existing export controls, or current registration status, resting its case entirely on the technology's Chinese origin and talent links.
The Manus enforcement suggests a new principle: technological nationality follows origin, not corporate domicile. China's Ministry of Commerce has effectively taken the stance that a technology's development location, not the company's registration, determines Beijing's regulatory authority. This has significant implications for Chinese-origin AI companies that have redomiciled offshore. Experts highlight that Singapore incorporation alone no longer de-risks deals from Chinese regulatory reach, introducing a 'reversibility risk' that complex deal structures cannot easily mitigate for Chinese-origin AI.
Following the Manus enforcement, China swiftly codified this principle with the Regulation on Outbound Investment (State Council Decree No. 837), effective July 1, 2026. This regulation is the first State Council-level administrative rule governing outbound investment. Article 13 explicitly prohibits exporting or transferring restricted technologies, services, or data, even through indirect means like relocating personnel or cross-border technical guidance. It introduces 'full-process supervision,' extending regulatory concern to an investment's entire lifecycle and establishing an outbound investment security review mechanism to counter inbound FISR.
U.S. acquirers of Chinese-origin AI now face simultaneous regulatory jeopardy from both the U.S. and Chinese governments. On the U.S. side, the Outbound Investment Rule (Executive Order 14105) restricts investments in AI-related entities involving Chinese persons. Concurrently, China's Catalogue of Technologies Prohibited or Restricted from Export covers core AI techniques. This creates 'dual jeopardy,' where companies could violate U.S. laws for investing in Chinese-origin AI and Chinese laws for acquiring such technology without approval, demonstrating that both jurisdictions may have grounds to intervene in a transaction.
In-house counsel must adopt a new diligence framework. This involves tracing technological nationality in every AI deal, identifying where R&D was performed, the technical team's training history, and whether IP transfers or restructurings received Chinese approval. Counsel should assume post-closing regulatory risk for Chinese-origin AI, accounting for potential unwind orders in deal structures. Dual-jurisdiction exposure must be mapped at the term sheet stage. Furthermore, existing AI investments and vendor relationships should be audited for Chinese-origin technology or talent, and a 'China regulatory discount' should be factored into valuations for cross-border exits of Chinese-origin AI ventures to U.S. buyers.
The Manus unwind is a watershed moment, establishing a new regulatory baseline where Beijing asserts authority over technology considered Chinese in origin, irrespective of offshore structures or corporate reorganizations. This authority is now statutory, prospective, and enforceable throughout an investment's lifecycle, reinforced by the new State Council regulation. For legal teams, the key takeaway is to look beyond corporate formalities to the fundamental origins of the technology, talent, and data being acquired in any AI-related transaction. The question is no longer where the company is registered, but where the technology was conceived and developed.