The Meta-Manus deal: what China’s foreign investment security review means for cross-border M&A in sensitive sectors
Kenneth Zhou
JunHe, Beijing
Executive summary
On 27 April 2026, the office of the working mechanism for foreign investment security review under China’s National Development and Reform Commission (NDRC) prohibited Meta’s proposed acquisition of Manus, a Chinese artificial intelligence (AI) agent company operated by Butterfly Effect, and required the parties to unwind the transaction arrangements.[1] This is likely the first publicly reported prohibition of a foreign acquisition in China’s AI sector under China’s Foreign Investment Security Review Measures,[2] which took effect in 2021.
The reported decision is significant not only because of the size and profile of the transaction, but also because it appears to confirm that China’s foreign investment security review regime will be applied aggressively to AI-related transactions involving key technologies, data capabilities and core technical talent. The case also suggests that offshore restructuring, relocation of headquarters or reduction of onshore operations may not be sufficient to remove a transaction from the scope of China’s national security review where the underlying technology, personnel, data or business substance remains closely connected to China.
This development is particularly important against the backdrop of the current US-China geopolitical landscape, in which strategic competition increasingly centres on control over critical, emerging technologies and supply chain security. Areas such as AI, robotics, semiconductors, advanced manufacturing, quantum computing and other frontier technologies are now viewed not only as drivers of commercial growth, but also as core components of national security, industrial policy and long-term technological leadership.
As a result, cross-border investments and acquisitions in these sectors are more likely to attract heightened scrutiny from regulators in both China and the US. Transaction parties should expect national security considerations to play a central role in deal planning, structuring and execution.
Background: the Meta-Manus transaction
In December 2025, Meta announced its proposed acquisition of Manus, a high-profile AI agent company founded by a Chinese team and widely regarded as a leading player in the AI agent sector, at a valuation of more than US$2bn. The transaction contemplated not only the acquisition of the target business, but also the integration of Manus’s core technical talent and underlying technology.
Manus had previously undertaken a series of restructuring steps, including relocating its headquarters from China to Singapore, reducing its domestic workforce and shifting its commercial focus toward overseas markets. This reflects a common pattern in China’s technology sector, where startups establish offshore corporate structures to access global capital, operate within a more flexible regulatory environment with respect to export controls, data movement and investment restrictions, while continuing to retain access to technological talent, data and the vast market in China.
Since its announcement, the deal has attracted significant public attention, not only because Manus’s founders chose to shut down the company’s China operations and relocate to Singapore, but also, and more importantly, because Manus is viewed as a leading Chinese AI company being acquired by a major US technology company amid the highly dynamic and sensitive US-China competition for technological leadership.
As a result, the transaction has drawn close regulatory scrutiny in China. For purposes of foreign investment security review, Chinese regulators generally focus on the substance of a transaction rather than the formal location of the holding company. Where core technology was developed in China, key personnel remain closely connected to China, or relevant data, R&D and commercial operations retain a substantial onshore nexus, the transaction may still be viewed as involving foreign investment in a sensitive Chinese business, even if part of the corporate structure has been moved offshore.
The regulatory risk also extends beyond foreign investment security review. A transaction of this nature may also raise issues under China’s technology export control regime, particularly if it involves sophisticated AI algorithms or other advanced software technologies. In recent years, China has tightened its controls over the export of sensitive technologies, data, products and services. Depending on the nature of the technology involved, regulators may scrutinise whether core AI models, algorithms, source code, training methods or related technical know-how may be transferred offshore only with prior approval.
Overview of China’s Foreign Investment Security Review Measures
China’s Foreign Investment Security Review Measures (the ‘Measures’) apply to direct and indirect foreign investments in China, including greenfield investments, acquisitions of equity interests or assets of domestic enterprises, and investments carried out through other means. The day-to-day administration of the review regime is handled by the office of the working mechanism housed at the NDRC, with the NDRC and the Ministry of Commerce (MOFCOM) jointly leading the mechanism.
Under the Measures, relevant parties must make a filing before implementing an investment if the transaction falls within certain sensitive categories. These include, first, investments in military and defence-related sectors, military-support sectors or areas surrounding military facilities, and second, investments in important sectors related to national security where the foreign investor will obtain actual control over the target. The listed sectors include important agricultural products, important energy and resources, major equipment manufacturing, important infrastructure, important transport services, important cultural products and services, important information technology and internet products and services, important financial services, key technologies and other important fields.
‘Control’ is defined broadly. It includes holding more than 50 per cent of the target’s equity, holding less than 50 per cent but having voting rights sufficient to exert material influence over shareholder or board decisions, or otherwise having the ability to materially influence the target’s business decisions, personnel, finance or technology.
The review process generally consists of three stages: a preliminary review of 15 business days, a general review of 30 business days and a special review of 60 business days, which may be extended in special circumstances.
The possible outcomes are clearance, conditional clearance or prohibition. If a transaction is prohibited, the authorities may require the parties to terminate the transaction, dispose of equity or assets, unwind the transaction or restore the parties to their pre-transaction position. Failure to file, filing false or misleading information or failing to comply with a review decision may result in additional regulatory consequences, including credit-related sanctions.
The Measures also extend to a foreign investor’s acquisition of shares in a domestic enterprise through the public securities markets. In particular, where a foreign investor acquires shares of a domestic enterprise through a stock exchange and the transaction implicates national security, the transaction may also be subject to national security review in China.
The two key agencies administering the review, NDRC and MOFCOM, are two of China’s most influential economic regulatory agencies. NDRC plays a central role in macroeconomic planning, industrial policy and strategic investment oversight, while MOFCOM is a key authority on foreign investment, trade and commercial regulation.
Comparison with US CFIUS: cross-border M&A in sensitive sectors now faces dual national security scrutiny
China’s Foreign Investment Security Review Measures are broadly comparable in function to the Committee on Foreign Investment in the United States (CFIUS) in the US. Both regimes operate outside ordinary merger control and general foreign investment approval processes; both grant regulators broad discretion to review transactions that may affect national security; and both permit regulators to clear, place conditions on, or prohibit a transaction. In each case, the analysis turns less on formal deal structure than on the substance of the transaction, including:
- the nature of the target business;
- the sensitivity of the technology;
- the connection to critical sectors or infrastructure;
- the degree of foreign influence or control; and
- the broader political, economic and strategic implications of the deal.
The CFIUS regime has undergone significant expansion in recent years. The CFIUS has long reviewed certain foreign investments in US businesses, but its jurisdiction broadened materially following the enactment of the Foreign Investment Risk Review Modernization Act (FIRRMA) in 2018. The FIRRMA extended the CFIUS review beyond traditional control transactions to cover certain non-controlling investments in US businesses involved with critical technology, critical infrastructure or sensitive personal data – commonly referred to as TID. It also strengthened the authority of the CFIUS to review transactions involving substantial foreign government interests and introduced mandatory filing requirements for certain categories of deals. During the Trump administration, the CFIUS became markedly more assertive, particularly with respect to Chinese investment in advanced technologies and data-driven businesses, reflecting a broader shift toward treating economic security and technological competition as national security priorities.
In parallel, the US has in recent years developed a separate outbound investment screening regime – often described as ‘reverse CFIUS’ (the final rule published the Treasury Department took effect in January 2025) – designed to restrict certain US investments into countries of concern, including China, in sensitive sectors such as semiconductors and microelectronics, quantum information technologies, and certain AI systems. Although this outbound regime is narrower than the CFIUS in formal scope, it reflects a significant policy development: US national security scrutiny now extends not only to foreign capital entering the US, but also to US capital, expertise and know-how flowing to sensitive sectors linked to geopolitical competitors.
Against this backdrop, the Manus deal appears exposed on both sides of the Pacific. On the Chinese side, regulators may view the transaction as involving the foreign acquisition of a strategically significant Chinese AI business, including advanced technology, data and highly skilled talent, notwithstanding the relocation of part of the corporate structure to Singapore. On the US side, the transaction also sits within a far more restrictive policy environment for China-related technology deals. Even if the transaction structure does not fit neatly within the traditional scope of inbound CFIUS review, the broader US framework now reflects heightened concern over transactions that could facilitate the transfer, development or strategic enhancement of sensitive technologies involving China, whether through inbound investment, outbound investment or associated transfers of technology and know-how.
Why China blocked the Meta-Manus deal
The Chinese government’s decision to block Meta’s acquisition of Manus appears to have been driven by the convergence of several reinforcing concerns: the transfer of strategically significant AI technology, the potential offshore movement of large volumes of China-linked data and highly skilled talent, and an apparent attempt to evade Chinese regulation through an offshore restructuring. More broadly, the transaction may have been viewed as the acquisition of a leading Chinese AI company by a major US technology company in circumstances that could strengthen US technological competitiveness at the expense of China’s own strategic position. In that sense, the deal was not merely a corporate acquisition, but a proposed transfer of core Chinese technological assets that directly implicated China’s national security and industrial policy priorities.
First, the deal appears to have raised serious concerns under China’s technology export control regime. Manus’s core value reportedly lies in its advanced AI agent technology, including algorithms, system architecture and multi-model orchestration capabilities developed substantially in China. To the extent those technologies fall within categories of restricted or controlled technologies, their transfer to a foreign acquirer could not lawfully proceed without prior regulatory approval. From the perspective of Chinese authorities, that concern would not be resolved merely because part of the corporate structure had been moved to Singapore.
Second, the transaction likely triggered acute data security concerns. Manus’ AI systems reportedly depended on significant volumes of user, operational and training data/models connected to China. Where such data may include sensitive personal information, commercially important data or other categories of ‘important data’ under Chinese law, any transfer or offshore access arrangement would raise immediate regulatory issues. Chinese regulators have in recent years treated data not merely as a commercial asset, but as a matter of sovereignty, security and strategic control.
Third, the authorities likely viewed Manus’ move to Singapore as a restructuring that did not alter the Chinese character of the underlying business. The key regulatory principle here is substance over form. Even if Manus shifted its holding structure offshore, Chinese regulators could still look through that structure and conclude that the company’s core technology, founding team, engineering capabilities, historical data accumulation and commercial roots remained closely tied to China. In that respect, the Manus case sends a clear message that a ‘Singapore-relocation’ strategy will not succeed where the underlying business remains, in substance, a Chinese strategic asset.
Fourth, and perhaps most importantly, China has become increasingly determined to protect its position in advanced technology sectors, and AI – particularly higher-level agentic systems – is now viewed as a strategic capability with long-term economic, industrial and geopolitical significance. Manus may therefore have been seen not simply as a startup acquisition target, but as a representative domestic AI platform whose transfer to a major US acquirer would risk weakening China’s competitive position in a strategically important field, particularly in the context of intensifying US-China competition for technological leadership.
For those reasons, the decision to block the transaction can be understood as more than a narrow, deal-specific ruling. It signals that Chinese authorities are prepared to approach AI-related and other sensitive-sector cross-border acquisitions with a high degree of caution.
The broader message to the market is clear: where a business’s core value was created through Chinese technology, Chinese talent and China-linked data, a later offshore restructuring will not necessarily insulate a sale to a foreign buyer from Chinese national security, export control or data governance review. Those risks are likely to be even more pronounced where the proposed acquirer is a major US company.
Key takeaways for companies, investors and deal advisers
The Manus case underscores that ‘control’ investments in AI and other key technology sectors require national security analysis at the outset of a transaction. Where a foreign investor seeks to acquire control over a Chinese business involved in AI models, AI agents, semiconductors, advanced software, computing infrastructure or other sensitive technologies, parties should assume that foreign investment security review may be relevant and should assess filing risk before signing, and certainly before closing.
The case also makes clear that offshore restructuring does not reliably remove a transaction from Chinese national security jurisdiction. Relocating a parent company, reducing the domestic workforce or transferring technology and personnel offshore will not necessarily avoid review if the underlying technology, data, personnel or operations remain materially connected to China. In practice, Chinese regulators are likely to look through form to substance.
Foreign investment security review is only one part of the broader regulatory framework. Even if a transaction can be filed or cleared under the Measures, separate issues may arise under China’s technology export control rules, merger control regime, data compliance framework, cybersecurity review regime and other applicable regulatory requirements. AI transactions, in particular, often raise several of these issues at once, requiring a coordinated regulatory assessment rather than a narrow focus on any single approval regime.
The Manus case also highlights the need for investors and founders to adopt a realistic regulatory strategy from the outset. Investors and founders should have a clear understanding of the geopolitical landscape and its implications for future exit constraints, transfers of control rights, intellectual property ownership, R&D location and key personnel mobility. It is no longer realistic to assume that a high-value exit to an offshore technology buyer will remain available regardless of the sector’s strategic sensitivity.
In other words, the long-held assumption that ‘technology is borderless’ no longer holds. Today, technology is increasingly bounded by national borders, regulatory regimes and geopolitical competition.
Where a transaction involves highly sensitive technology, significant control rights or an apparent attempt to evade Chinese review through pre-transaction restructuring, the risk of prohibition increases materially. Earlier landmark cases under China’s foreign investment review framework, although predating the current Measures, including the Coca-Cola/Huiyuan and Carlyle/Xugong matters, reflected similar regulatory instincts in sectors regarded as strategically important. The Manus case would indicate that this same policy logic now applies with equal force to frontier AI technologies. That risk increases substantially where the transaction also implicates US-China competition in technology and other strategic sectors.
Conclusion
The real significance of the Manus case lies not only in the fate of a single transaction, but in what it reveals about the new political geography of technology. AI may no longer be treated as a neutral commercial asset that can be freely developed in one country, financed through another, and sold on a global market. Governments increasingly view it as a strategic resource tied to national power, economic resilience, data sovereignty and long-term technological leadership. In this environment, the lines between corporate dealmaking, industrial policy and geopolitical rivalry are becoming increasingly blurred.
The Manus decision makes clear that when a technology company is seen as part of a nation’s strategic future, questions of ownership, control and transfer become matters of national security, not just matters of contract. For global technology companies and investors, the central challenge is no longer simply how to structure cross-border deals, but whether some deals will remain politically possible at all in an era defined by US-China competition for technological leadership.