DoorDash, Siemens and Airbnb Turn to Chinese AI Models to Cut Costs

A quieter shift in the artificial-intelligence market may prove as consequential as any single product launch. The Financial Times reported that companies including the food-delivery firm DoorDash, the German industrial group Siemens and the lodging platform Airbnb are turning to Chinese-developed AI models to curb rising costs and reduce their reliance on United States technology.

The driver is price. As businesses move from experimenting with AI to running it across large workloads, the cost of the leading American models has become a material expense. Chinese models that deliver comparable performance at a lower price reduce that cost, and for firms outside the United States they also reduce dependence on a small number of American suppliers whose access could be constrained by policy.

The development cuts against a common assumption that US export controls would keep Chinese AI confined to China. Instead, restrictions intended to slow China's chip access have pushed its developers to build efficient models and offer them abroad, giving global buyers a second source. That is the mechanism by which a parallel technology stack spreads beyond its home market.

For American AI firms, the competitive threat is to pricing power rather than to their position as the highest-performing. If capable models become a low-cost commodity, the premium that leaders can charge erodes, and the economics that justify enormous data-center investment come under pressure.