In December 2025, Meta acquired Manus, a Chinese AI firm, for US$2 billion. It was one of the first major acquisitions of a Chinese artificial intelligence company by an American tech giant. Such deals are rare due to US regulatory restrictions and stark valuation disparities between Chinese and Western AI markets. What made the deal possible? Manus had redomiciled to Singapore six months earlier, rebranding itself as Singaporean to sidestep US restrictions on Chinese tech. Within days of the announcement, Beijing launched a regulatory investigation, raising questions about whether this escape route will remain open for companies hoping to replicate Manus's success.
The Manus deal exposed a structural problem facing Chinese AI startups: severe capital constraints. Zhipu and MiniMax, two of China’s six “AI tigers” (the elite cohort of independent large language model developers), recently floated on the Hong Kong Stock Exchange at valuations below US$7 billion each. Compare that to their American counterparts: OpenAI at US$500 billion and Anthropic at US$183 billion. For Chinese entrepreneurs building cutting-edge AI, domestic capital markets offer only a fraction of what Western funding can provide. This is a function of both China's conservative public markets and the concentration of global AI investment capital in Silicon Valley.
Singapore has systematically capitalised on rising US–China tensions and Hong Kong’s decline as Asia’s financial hub.
This valuation gap is driving “China-shedding”: Chinese-born companies relocating to Singapore, severing mainland ties, and repositioning as "Singaporean" to access Western capital while avoiding geopolitical restrictions.
Manus followed this playbook precisely. Anticipating a Silicon Valley exit, the company moved its headquarters from Beijing to Singapore in 2024. American venture capital firm Benchmark led a US$75 million funding round, triggering a restructuring: Manus shrank its Shanghai research teams and relocated its C-suite to Singapore, rebranding itself enough to close a deal with Meta.
US outbound investment rules restrict American acquisitions of Chinese AI firms, while China's equivalent to the Committee on Foreign Investment scrutinises technology transfers. This creates a two-way regulatory squeeze that makes direct US–China AI deals nearly impossible.
Singapore has become the perfect conduit for this strategy. The city-state offers a 17% corporate tax rate, robust funding ecosystem, and white-glove government services. But for Chinese companies specifically, Singapore has unique advantages: a Chinese-majority population where Mandarin is widely spoken alongside English, proximity to mainland networks, yet sufficient legal distance to bypass Western restrictions.
This positioning isn't accidental. Singapore has systematically capitalised on rising US–China tensions and Hong Kong’s decline as Asia’s financial hub. Post-Covid decoupling and Hong Kong's National Security Law accelerated the trend, with Western companies shifting Asia-Pacific headquarters to Singapore. Now Chinese tech firms are following the same migration path, but in reverse, seeking Western capital rather than Asian operations.
Beijing's investigation signals scrutiny of this practice, but full decoupling seems unlikely. Instead, expect messy restructurings that maintain plausible separation.
The venture capital industry offers a preview of how this might work. In 2023, Sequoia Capital executed such a plan, spinning off its entire China business and rebranding it as HongShan (the literal Mandarin translation of "Sequoia"). The Sequoia team in China either dispersed or relocated to Singapore, maintaining distance from both Beijing and Washington while preserving regional deal flow.
Singapore enforces Beijing’s technology controls and Washington’s compliance protocols, bending to both sides’ contradictory demands as long as deals keep flowing.
The HongShan split offers a template for how China-shedding might evolve. Rather than shutting down cross-border AI deals entirely, Beijing may force cleaner separations: Chinese founders keep mainland research and development operations under one entity while spinning off international subsidiaries that can pursue Western exits.
Despite the increasingly regulated flow of technology and capital, Singapore is poised for continued success because it is ruthlessly pragmatic. In the China-shedding model, Singapore becomes not just a relocation site for headquarters, but a genuine corporate bifurcation point, allowing Chinese innovation to access global capital through structurally independent vehicles. As a result, Singapore is constructing a role that neither Washington nor Beijing can easily replicate or eliminate.
Chinese AI companies need Singapore's legal infrastructure and Western-friendly brand to access global markets. American tech giants need Singapore's proximity to Asian talent and its ability to transact where direct US–China deals have become politically toxic. For a city-state of just six million people, facilitating even a fraction of global AI capital flows represents an enormous economic multiplier.
But pragmatism demands flexibility. Singapore enforces Beijing’s technology controls and Washington’s compliance protocols, bending to both sides’ contradictory demands as long as deals keep flowing. The Manus investigation tests whether this tightrope walk is sustainable: Can Singapore satisfy both superpowers’ security concerns while remaining the essential conduit between them?
The alternative is stark. If Beijing forces Chinese AI innovation to remain fully domestic, Chinese startups face permanent capital constraints. If Washington blocks all China-adjacent AI deals, American firms lose access to innovative Chinese technologies that could accelerate their own development. Both outcomes leave money on the table and create an opening for a pragmatic intermediary.
Singapore is betting that neither superpower can afford to let that happen. But this strategy carries inherent risk: pragmatism works only as long as both sides value Singapore’s services more than they resent its refusal to pick a team. The Manus investigation suggests Beijing’s tolerance may be wearing thin. Washington’s could follow. The question isn’t whether Singapore can remain neutral. It’s whether it can remain indispensable.
