China is signaling that AI startups hoping to follow Meta’s playbook may need to slow down—and think twice.
On Thursday, Beijing said it is reviewing Meta’s acquisition of AI startup Manus, a move that underscores how seriously China now treats advanced AI as a strategic national asset. The message is clear: global deals are still possible, but they won’t fly under the radar anymore.
A Deal That Got Beijing’s Attention
Meta acquired Manus last month as part of its broader push to scale AI agents across consumer and enterprise products. While neither company disclosed financial details, reports from The Wall Street Journal suggest the transaction topped $2 billion.
That price tag—and Manus’ origins—appear to be what caught regulators’ attention.
Manus began inside a Chinese startup known as Butterfly Effect before spinning out and relocating to Singapore earlier this year. Since then, it has grown at breakneck speed, launching an AI agent capable of handling tasks like research, coding, and data analysis.
Why China Is Stepping In
China’s Ministry of Commerce said it will assess whether the acquisition complies with export control rules and regulations governing technology transfers and overseas investments.
In plain terms: Beijing wants to know how much of Manus’ underlying AI technology was developed in China—and where that know-how might end up.
At a press briefing, ministry spokesperson He Yadong emphasized that China supports cross-border tech cooperation, as long as it follows domestic laws. That caveat is doing a lot of work.
AI Agents Are Now “Strategic”
Industry analysts say this isn’t about blocking the deal outright. It’s about leverage.
Advanced AI agents—systems that can autonomously perform complex tasks—are increasingly seen by governments as dual-use technologies with economic and national security implications. Letting them slip overseas without oversight is no longer acceptable in Beijing’s view.
The likely outcome, experts say, is a slower approval process and possible conditions on how Manus’ technology can be used or commercialized outside China.
Meta’s Bigger AI Bet
The scrutiny comes as Meta pours money into AI to keep pace with rivals like OpenAI and Google.
Beyond Manus, Meta recently took a major stake in Scale AI and has been reorganizing internally to prioritize product-focused generative AI. CEO Mark Zuckerberg has made it clear that AI agents will play a central role in Meta’s future platforms.
What This Means for Startups
For Chinese-founded startups—or those built on China-developed IP—the takeaway is stark.
Fast overseas exits, especially to U.S. tech giants, are no longer just business decisions. They’re geopolitical events.
Founders may need to factor in longer timelines, regulatory uncertainty, and potential limits on how their technology can be deployed globally. Some may even rethink where they incorporate or move key operations much earlier in their lifecycle.
The Bigger Picture
China isn’t shutting the door on global AI deals. But it is installing a lock—and keeping the key close.
As AI becomes more powerful and more political, the era of frictionless cross-border acquisitions is fading fast.
In the global AI race, innovation still moves quickly. Permission doesn’t.