
Tech • IA • Crypto
China has blocked Meta’s $2 billion acquisition of AI firm Manus, signaling a major geopolitical escalation in control over strategic artificial intelligence technologies.
On April 27, China’s National Development and Reform Commission (NDRC) ordered the reversal of Meta’s acquisition of Manus, despite the deal already being finalized. The transaction, worth around $2 billion, included team integration, data transfer, and operational consolidation. Authorities demanded a full rollback, including funds, intellectual property, and personnel.
The decision relies on China’s foreign investment security mechanism, introduced in 2021 but rarely enforced until now. Officials argued the deal involved sensitive technologies and improperly transferred strategic assets abroad without approval, marking an aggressive expansion of regulatory reach.
Manus is not a traditional language model but an agentic orchestration platform, capable of autonomously coordinating multi-step AI tasks across systems. The company reportedly reached $100 million in monthly revenue within eight months, making it a high-value strategic asset in the rapidly growing field of AI agents.
Chinese authorities cited unauthorized export of training data and behavioral data from Chinese users as a core issue. They argue these datasets underpin Manus’s capabilities and cannot legally be transferred to foreign entities, especially major U.S. tech firms like Meta.
Meta has contested the decision, stating it complied with international regulations. However, China has enforced compliance by summoning Manus’s founders—who have not been allowed to leave the country—and demanding immediate reversal. The dispute is expected to surface in upcoming high-level talks between Xi Jinping and Donald Trump.
The move suggests China is willing to assert control not only over domestic companies but also over foreign acquisitions involving Chinese-origin technology. This could reshape global AI investment flows and accelerate a technological decoupling between China and the United States.
Meanwhile, French company Mistral AI has launched Mistral Workflows, a platform designed for large-scale enterprise AI automation. Targeting firms like CMA CGM, La Banque Postale, and France Travail, it focuses on reliability, monitoring, and orchestration of complex AI processes across organizations.
In parallel developments, OpenAI has expanded beyond its historical reliance on Microsoft, making its models available on Amazon Web Services (AWS). This marks a broader shift toward multicloud AI ecosystems, increasing competition and accessibility across providers.
Industry signals indicate OpenAI, in collaboration with former Apple designer Jony Ive, is developing a new AI-centric hardware device. Early reports suggest a custom chip designed by Qualcomm and MediaTek, with potential commercialization by 2028, aiming to replace traditional app-based smartphones with fully agent-driven interfaces.
The forced reversal of Meta’s Manus acquisition highlights how artificial intelligence has become a central axis of geopolitical power, with nations increasingly asserting control over data, talent, and technology in a rapidly fragmenting global landscape.