Regulatory  ·  2026-04-28

China blocks Meta's $2 billion acquisition of agentic AI startup Manus

RegulatoryHigh impactChina
China's National Development and Reform Commission (NDRC) on 27 April 2026 ordered Meta and Manus to unwind Meta's $2–3 billion acquisition of the Singapore-based agentic AI startup. Manus, founded by Chinese engineers and relocated from Beijing to Singapore in mid-2025, had already integrated ~100 employees into Meta's Singapore offices and appointed CEO Xiao Hong to report directly to Meta COO Javier Olivan. The NDRC statement cited compliance with laws and regulations but provided no detailed explanation.
This represents one of China's most significant interventions in a cross-border AI deal and sets a precedent for how Beijing will treat AI technology transfers, particularly involving agentic AI capabilities. The move comes weeks before a US-China summit and extends beyond bilateral tensions into a broader assertion of control over frontier AI development originating in China. For AI security practitioners, this signals that agentic AI systems—capable of autonomous task execution—are now categorically treated as strategic assets subject to export control-style scrutiny, even when companies relocate. The practical impact is immediate: Manus founders are reportedly under exit bans, and the deal must be unwound despite operational integration already underway.
Organizations with cross-border AI partnerships, especially involving agentic or autonomous systems, should review jurisdiction and ownership structures now. If your firm is acquiring or partnering with AI startups that have Chinese founders, Chinese R&D centers, or were previously domiciled in China, expect heightened regulatory review from both Chinese and US authorities. Update M&A due diligence checklists to include country-of-origin analysis for founding teams and initial intellectual property development, not just current corporate domicile. For clients operating in or near China, this ruling underscores that agentic AI is now a regulated technology class—consider whether your deployment plans require advance regulatory consultation.
Sources
TechCrunch (primary)CNNBloomberg
See this in the live feed Explore related AI security and governance findings — updated every morning.
Open the feed →