Policy

Meta Cedes Control of Chinese AI Startup Manus

Meta has relinquished control of Chinese artificial intelligence startup Manus after China blocked the acquisition, highlighting the growing regulatory hurdles facing cross-border AI deals.

Computerworld AI4 days agoPolicy
Image: Computerworld AI

Chinese artificial intelligence startup Manus has announced plans to operate as an independent entity once again, following the reversal of its acquisition by Meta. The US social media giant had agreed to purchase Manus last year, but the deal quickly drew scrutiny from Chinese regulators concerned about export control violations. In April, China's National Development and Reform Commission officially blocked the transaction, ending Meta's eight-month control of the startup.

As Manus transitions back to independence, the startup must comply with strict regulatory mandates. This includes deleting specific user data that was collected during the eight months it operated under Meta's umbrella. The unwinding of this deal highlights the immediate operational and data-handling challenges that AI practitioners and startups face when cross-border mergers collapse under regulatory pressure.

The collapsed deal is part of a broader geopolitical struggle between the US and China over AI supremacy. Both nations are increasingly protective of their domestic technology. In March, the US administration introduced a new cybersecurity policy following a suspected Chinese cyber-attack on the FBI. That same month, the US-China Economic and Security Review Commission warned that China's utilization of open-source AI tools could grant it an unregulatable economic advantage. These events follow a series of restrictions placed on Nvidia chip exports last year.

For AI developers, enterprise buyers, and investors, this failed acquisition serves as a warning. Practitioners must prepare for heightened scrutiny when dealing with international vendors or cross-border data flows. Navigating the fragmented regulatory landscape means that technical architectures must be designed with data portability and sovereignty in mind, as sudden regulatory interventions can force companies to purge valuable datasets or completely restructure their corporate alignments.

This is our own summary of reporting by Computerworld AI

More in Policy