The global AI race is no longer just about building faster models, but about controlling where that intelligence lives.

China’s National Development and Reform Commission has ordered Meta to reverse its USD 2.5 billion acquisition of Manus, citing national security concerns tied to the transfer of AI technology.

Why You Should Care

China’s decision to force Meta to unwind its USD 2.5 billion acquisition of AI startup Manus signals a sharper shift in how advanced technology is being governed.

For founders, investors, and operators, the message is clear: in frontier sectors like AI, ownership, talent movement, and data flows are no longer purely commercial decisions. They are geopolitical ones.

As cross-border deals face deeper scrutiny, scaling globally may increasingly depend not just on capital or capability, but on regulatory alignment across markets.

The Details

The Wall Street Journal reports that China’s National Development and Reform Commission has ordered Meta to reverse its acquisition of Manus, citing national security concerns tied to the transfer of AI technology.

The deal, completed in late December, valued Manus at over USD 2 billion. This brought an AI agent capable of executing complex tasks, such as generating research reports and preparing presentations, into Meta.

Manus traces its origins to Beijing Butterfly Effect Technology, founded in 2022. While the company later shifted operations and staff to Singapore following foreign investment, Chinese regulators maintained that the core entity remained subject to domestic oversight.

Authorities launched a formal review shortly after the acquisition and have now given Meta a matter of weeks to reverse the transaction. This includes restoring Manus’s Chinese assets and removing any transferred data or technology.

The scrutiny reflects concern within Beijing that similar structures could allow Chinese-developed technologies to move offshore without approval. Reports indicate that Manus’s founders were questioned during the investigation and subjected to travel restrictions during the review period.

Meta has stated that the transaction complied with applicable laws and that it expects a resolution, though preparations to unwind the deal are already underway.

The Ripple

The implications extend well beyond a single transaction.

Cross-border AI deals are entering a different phase of scrutiny, especially when the underlying technology originates in China.

The Manus case shows that relocating a company or restructuring ownership does not necessarily remove regulatory oversight. Even after a deal closes, authorities can still intervene when they determine that core technology, talent, or intellectual property serves domestic interests.

It also introduces a new layer of uncertainty around acquisitions. Completed deals are no longer final by default. This is especially clear in sectors like AI, where governments consider data and know-how as strategic assets. The ability to transfer those assets across borders is becoming less predictable.

Company structure now matters more than ever. Where teams are based, where technology is developed, and how entities are structured across jurisdictions could directly influence whether an acquisition can proceed or be held up over time.

What to Watch

How fully and quickly the deal reverses will offer early signals on how enforceable these restrictions are in practice.

More broadly, any expansion of similar reviews, or stricter rules around talent, data, and IP transfers, will indicate how far governments are willing to go in localising control over AI development.

The direction is becoming clearer: as AI capabilities scale, so does the importance of where they are built, held, and governed.