The Chinese government has formally blocked Meta’s acquisition of Manus AI, citing national security concerns and shutting down what would have been one of the largest cross-border AI deals of 2026. Manus is one of the most prominent autonomous agent startups to come out of China, and Meta had publicly framed the acquisition as a way to accelerate its agent roadmap with technology that has already proven itself in Chinese enterprise deployments. The Chinese decision lands at a delicate moment for US-China technology relations and signals how Beijing intends to handle frontier AI assets going forward.
Manus AI launched in 2025 with an autonomous agent platform that gained traction quickly among Chinese e-commerce, logistics, and financial services companies. The product line specializes in long-running multi-step agents that can execute business processes spanning hours or days, with checkpoints for human review at decision boundaries. By early 2026, Manus had reportedly signed enterprise customers including Alibaba subsidiaries, Meituan, and several state-owned banks. That customer concentration, combined with the fact that Manus’s training data includes proprietary Chinese business workflows, made the company strategically valuable in ways that pure model labs are not.
Meta’s interest was reportedly driven by the gap between its own agent capabilities and what Anthropic and OpenAI have shipped. Meta’s Llama models compete on raw benchmarks, but the company has been slower to ship production-ready agent infrastructure for enterprise customers. Acquiring Manus would have given Meta a working agent stack, an experienced engineering team, and a foothold in the Chinese enterprise market in a single transaction. The reported valuation discussed informally was in the multi-billion-dollar range.
The Chinese rationale, as cited in official statements, focuses on national security and the risk of sensitive Chinese business data flowing to a US technology company. The Cyberspace Administration of China and the State Administration for Market Regulation have both been increasingly active in reviewing foreign acquisitions of Chinese AI companies, and the Manus block sits within a broader pattern that includes earlier reviews of cross-border deals involving facial recognition, autonomous driving, and defense AI vendors.
The geopolitical context matters. The deal block comes as US-China technology decoupling continues to intensify, with successive US administrations expanding export controls on advanced chips, EUV equipment, and AI training compute. China’s response has been to tighten domestic ownership of strategic technology assets, and Manus is squarely within that strategic category. The message to other Chinese AI startups is clear: foreign acquisition is no longer a viable exit path for companies with sensitive customer data or strategically important technology.
For Meta, the decision is a setback but not a fatal one. The company has substantial internal AI resources and can continue to build agent infrastructure organically. The bigger question is whether Meta will pivot to acquiring agent startups in jurisdictions outside China, or whether the company will accelerate its open-source Llama strategy as a way to recruit external developers into its ecosystem.
For sales teams tired of cold leads, slow customer responses, and manual processes, Dapta is the ultimate tool.
Dapta is the leading platform for creating AI sales agents specifically designed to increase inbound lead conversion. Respond to your leads in less than a minute with voice AI and WhatsApp that converts.
If you want your team to sell more while AI handles the complex stuff, you have to try it.
For LATAM and global markets, the Manus block reinforces a pattern that has been emerging since 2024: the AI startup ecosystem is increasingly fragmented along geopolitical lines. Chinese AI companies will sell primarily to Chinese and emerging-market customers, US AI companies will dominate the OECD enterprise market, and LATAM buyers will increasingly need to evaluate whether their AI vendors carry geopolitical risk that affects long-term contract viability. Platforms operating across LATAM markets benefit from being regionally focused, since regulatory blowback on cross-border AI flows is unlikely to disrupt locally based operations.
For Manus specifically, the block likely accelerates the company’s IPO timeline on the Hong Kong or Shanghai exchanges, since liquidity for early investors will need to come from public markets rather than acquisition. For other Chinese AI startups in similar positions, the lesson is to plan for domestic exits and to invest heavily in domestic enterprise sales motion. The cross-border M&A path is effectively closed.
The bigger picture is that AI is now a geopolitically managed asset class. Governments around the world are deciding which AI capabilities they will allow to cross national borders, and Manus is the latest data point in a trend that will define how the next decade of AI investment unfolds.
