Quick Tips for Understanding the Decision
- Look beyond the headquarters: A company’s legal registration in Singapore does not erase its foundational ties to Chinese talent, technology, and data infrastructure.
- Understand tech export laws: Cross-border AI mergers now face intense scrutiny under national security regulations and technology export control mechanisms.
- Watch the talent pool: High-value tech acquisitions often function as “acqui-hires,” making the retention and cross-border transfer of core engineering teams a primary regulatory target.
The Rise of Manus and the Acquisition Timeline
To understand the sudden halt of this historic merger, we must examine the incredibly rapid rise of Manus. The startup’s journey from a local project to a global acquisition target happened in exactly one year.
From Beijing Startup to Global AI Sensation
In early 2024, founders Xiao Hong and Ji Yichao established Butterfly Effect, the parent company of Manus. The team initially based its operations in Beijing and Wuhan. On March 6, 2025, the company launched Manus, billing it as the world’s first general AI agent.
The product achieved viral success almost overnight. Users shared incredible demonstrations of the AI agent executing complex, multi-step tasks across social media platforms. The company utilized a strict waitlist and invite-code system, driving demand so high that invite codes sold for thousands of dollars on second-hand markets. Manus quickly earned the reputation of a massive breakthrough in the AI application layer.
The Strategic Move to Singapore
Despite relying heavily on Chinese engineering talent and infrastructure during its early development, Manus leadership had global ambitions. In April 2025, the company secured $75 million in funding led by Benchmark, a top-tier Silicon Valley venture capital firm. This investment pushed the startup’s valuation to $500 million.
Following this cash injection, Manus initiated a strategic pivot that many industry insiders described as “Singapore-washing.” The company actively distanced itself from its Chinese origins. It shut down domestic channels, cleared its Chinese social media accounts, and relocated its corporate headquarters and core team to Singapore. By May 2025, Manus opened its platform globally without invite codes, accelerating its international commercialization while restricting access within China.
Meta’s Multibillion-Dollar Move
By late 2025, the startup’s strategic pivot appeared highly successful. Manus announced its annual recurring revenue had crossed the $100 million mark. Concurrently, Meta CEO Mark Zuckerberg and CTO Andrew Bosworth were actively shifting Meta’s AI strategy from standard chatbots to highly capable, task-executing agents. Manus fit their exact needs.
In December 2025, Meta announced its intention to acquire Manus for approximately $2 billion. Financial disclosures later revealed the depth of the commitment. Meta’s fourth-quarter balance sheets showed an unusual $3.25 million increase in restricted cash, earmarked as an escrow deposit for a pending acquisition. Furthermore, the company recorded an accrued liability of $1.72 billion for acquisition obligations. The deal was incredibly close to completion.
The Regulatory Intervention
The massive acquisition immediately triggered regulatory alarms in China. Authorities launched a comprehensive review to evaluate whether the deal violated technology export controls. After a thorough assessment, Chinese officials issued a formal notice on April 27, 2026, prohibiting Meta from acquiring Manus and demanding the cancellation of the transaction. The $2 billion deal officially collapsed.
Legal and Geopolitical Considerations
The decision to block the merger sparked widespread international debate. Some critics accused China of overreaching, given that Manus operated out of Singapore at the time of the deal. However, examining the legal frameworks reveals a different reality.
Defining Jurisdiction and National Security
Critics often misuse the term “long-arm jurisdiction” to describe this regulatory block. Long-arm jurisdiction typically involves applying domestic laws to sanction purely third-party entities abroad. The Manus case does not fit this definition.
The regulatory focus was not on the company’s current mailing address, but on the origins of its intellectual property. Manus developed its early prototypes in China. It trained its foundational models using Chinese data and relied on domestic engineering talent. Under China’s Foreign Investment Security Review Measures and the Catalog of Technologies Prohibited or Restricted from Export, any cross-border transfer of such critical tech requires explicit government approval. China possessed a firm legal foundation to exercise jurisdiction over the underlying assets.
The New Normal of International M&A
Halting a major technology acquisition aligns perfectly with current international regulatory practices. Across the globe, cross-border mergers involving artificial intelligence, sensitive data, and core software face extreme scrutiny.
For example, in 2017, the Committee on Foreign Investment in the United States (CFIUS) blocked a Chinese firm from acquiring AppLovin, a US-based mobile marketing platform. The block occurred not because AppLovin possessed classified military tech, but because it controlled massive amounts of user behavior data. Governments worldwide now view digital data and advanced algorithms as critical sovereign assets. China’s intervention in the Manus deal mirrors these established global security protocols.
The “Acqui-Hire” Controversy
Another major factor in the regulatory block involved the nature of the acquisition itself. Legal analysts pointed out that Meta’s offer relied heavily on retaining the founding team. The deal included a massive $50 million retention bonus pool explicitly designed to lock in Xiao Hong and his core engineers.
Regulators increasingly view these transactions as “acqui-hires”—a corporate strategy used to bulk-buy top-tier talent rather than just software code. In the high-stakes AI race, elite engineers are incredibly scarce resources. Allowing a foreign tech giant to absorb a highly successful domestic engineering team presents clear risks to long-term national competitiveness.
Broader Impacts on AI Innovation and Global Collaboration
The collapse of the Manus acquisition represents a watershed moment for the global technology industry. It delivers clear lessons for venture capitalists, startup founders, and multinational corporations operating in the AI sector.
The End of Borderless Tech Startups
For decades, tech entrepreneurs operated under the assumption that software had no borders. A team could build a product in one country, register a holding company in another, and sell to a third. The Manus case shatters this illusion.
Technology startups must now navigate complex geopolitical realities from their inception. Attempting to bypass domestic regulations by suddenly shifting headquarters to a neutral hub like Singapore is no longer a viable strategy for companies holding sensitive intellectual property. Founders must build regulatory compliance into their business models from day one.
Implications for Foreign Investment
Some observers expressed concern that blocking the Meta deal signals a tightening of China’s business environment. However, restricting a highly sensitive AI merger does not contradict the country’s broader push for foreign investment.
China remains the world’s second-largest recipient of foreign direct investment. The government has consistently reduced its negative list, opening up sectors like high-end manufacturing and services. By clearly defining the boundaries of national security, regulators actually provide stability. Transparent rules allow international investors to operate with confidence in non-sensitive sectors, knowing exactly where the regulatory red lines exist.
What This Means for Future AI Deals
The failed Manus acquisition sets a permanent precedent for the artificial intelligence industry. Moving forward, cross-border M&A deals involving AI agents, large language models, or significant data sets will face intense friction.
Companies hoping to operate globally may need to explore dual-track development models. This involves building completely separate technical infrastructures and engineering teams for domestic and international markets. Alternatively, firms may pivot toward licensing agreements and minority stake investments rather than outright acquisitions. The era of frictionless, multibillion-dollar global tech buyouts has officially given way to an era of strict technological sovereignty.
Conclusion
The halting of the Manus acquisition proves that the rules of global technology expansion have fundamentally changed. Artificial intelligence is no longer just a commercial product; it is a critical pillar of national power. As governments worldwide tighten their grip on digital assets and engineering talent, tech companies must adapt to a landscape where innovation is inextricably linked to national security. Navigating this new reality requires deep legal foresight and a clear understanding of shifting geopolitical boundaries.