CHINA BUILDS AI FINANCING LOOP WITH STATE-BACKED BUYERS

The arrangement mirrors a financing model that has drawn criticism in the US around Nvidia's dominant position in AI infrastructure. State-backed training centers buy Unitree's robots, use them to generate data, then sell that data back to the company—creating a closed loop that concentrates capital and information within the ecosystem. This circular model raises questions about market efficiency and competition. Unlike traditional procurement, where independent buyers create diverse demand signals, state-backed entities buying machines specifically to feed data back to manufacturers essentially recycle capital within a controlled system. The pattern reflects how different markets are structuring AI development. While US regulators scrutinize tech giants' control over chip supply and training data, China's state-coordinated approach achieves similar consolidation through institutional buyers rather than corporate monopolies. Both models concentrate resources for AI advancement but through different mechanisms.