What China’s Manus decision means for the future of AI

By Avesta Afshari-Mehr

China has recently ordered Meta to unwind its $2billion acquisition of the AI app Manus: even though the deal has already been completed and partially integrated into Meta’s systems. It’s a rare and dramatic move, and it says a lot about where artificial intelligence sits in global politics right now. So what’s actually going on?

Manus isn’t just any AI application. Developed originally by a Chinese startup called Butterfly Effect, it focuses on building AI “agents”, tools that don’t just answer questions (like ChatGPT), but can carry out complex tasks independently, like organising files, running processes, and helping build software. This is why Meta wanted it. The company has been racing to keep up with rivals like OpenAI and Google, and buying Manus was a shortcut to getting ahead. The deal was announced in late 2025 and closed earlier this year, with Meta already starting to integrate the technology into its own tools.

But then China stepped in. Regulators, including powerful bodies like the National Development and Reform Commission, launched an investigation into whether the acquisition violated rules around foreign investment and technology transfer. Their conclusion: the deal must be undone completely, without question. But unwinding an acquisition at this stage isn’t like returning a faulty product. It could involve selling Manus back to its original investors, finding a new buyer, or restructuring ownership entirely. On top of that, Meta would have to stop using the underlying algorithms and potentially rebuild parts of its own systems. So why take such a drastic step?

The company has been racing to keep up with rivals like OpenAI and Google

At one level, this is about control. China has become increasingly protective of its tech sector, especially in areas like AI, which are seen as strategically critical. Allowing a major US company to absorb a promising AI tool raises concerns about losing technological advantage. And then there’s the bigger picture, in the sense that this decision comes at a time of growing tension between the US and China over technology. AI in particular has become a key battleground as both countries see it not just as an economic opportunity, but as something tied to national security and global influence.

Key sector analysts think the order to unwind the deal is less about reversing this specific acquisition and more about sending a message. If even a completed, high-profile deal can be challenged, it sets a precedent. Future companies might think twice before attempting similar cross-border acquisitions. But there are also real and serious consequences for the people involved. Several reports suggest that Chinese authorities had already restricted travel for some of Manus’s founders during the investigation. If Meta fails to comply, it could face penalties, limits on its business in China, or even legal action. Meanwhile, the US response has been predictably critical. Officials have framed the move as interference in normal business activity, reflecting broader concerns about how China regulates foreign companies.

It’s clear that the AI tools we use every day don’t exist in a vacuum. They’re shaped by global competition, government policy, corporate strategy, the works. A deal like this shows how quickly those forces can collide. And it also raises an interesting question for the future: can global tech companies still operate seamlessly across borders in an age where governments are becoming more protective? For now, Meta is stuck trying to untangle a deal it thought was complete. And Manus, once just another ambitious AI startup, has become a symbol of something much bigger.

Image Credit: Markus Spiske via Wikimedia Commons