China kills Meta’s acquisition of Manus as US-China AI rivalry deepens
Meta's AI ambitions hit a wall as Beijing orders the deal unwound over national security.
In a dramatic escalation of US-China AI tensions, China has formally blocked Meta's $2 billion acquisition of Manus, an AI startup founded by Chinese entrepreneurs Xiao Hong and Ji Yichao. The Chinese government ordered Meta to unwind the deal on April 27, citing national security concerns after months of investigation that began in January 2026. The founders had already relocated most of their team to Meta's Singapore office and cut ties with China, but authorities still intervened.
Manus burst onto the scene in March 2025 with its general AI agent, which uses Anthropic's Claude 3.7 Sonnet as a foundation and incorporates multiple sub-agents for planning, execution, and verification. Meta had begun integrating Manus into its Ads Manager platform as part of CEO Mark Zuckerberg's push for "personal AI superintelligence for everyone." The unwinding now threatens Manus's ability to use Anthropic's models, which are restricted in China, and represents a major setback to Meta's AI pivot after spending $80 billion on the metaverse. The case also signals that the "Singapore-washing" model—where Chinese founders relocate abroad—may no longer work, forcing them to set up outside China from day one.
- China formally ordered Meta to unwind its $2 billion acquisition of Manus on April 27, citing national security concerns after a months-long investigation.
- Manus uses multiple AI agents (planner, executor) built on Anthropic's Claude 3.7 Sonnet, and Meta had integrated it into Ads Manager.
- The deal's collapse threatens Manus's access to Anthropic's models and signals the end of the 'Singapore-washing' strategy for Chinese tech founders.
Why It Matters
This deal's collapse signals a new era where cross-border AI acquisitions are nearly impossible, reshaping global tech strategy.