China's technology sector is on the brink of a fundamental shift. Contrary to the current landscape, where hardware suppliers capture the bulk of margins in the field of artificial intelligence, the balance of power could change dramatically within two to three years. Internet platforms with massive user bases will begin to accumulate the lion's share of AI profits—but only if export restrictions on chips and infrastructure are relaxed.
My assessment of the situation aligns with that of leading analysts: the market is currently constrained by a shortage of computing power, which places "hardware" in a privileged position. However, once supply chains normalize, pricing power will inevitably shift to platforms that own data, audiences, and distribution channels. This is a classic monetization scenario that we have already observed in other technology cycles.
The Cost of the AI Race
The investment activity of Chinese giants speaks for itself. A telling example is Alibaba: capital expenditures nearly tripled in the second quarter, reaching 52.8 billion yuan ($7.86 billion). At the same time, the company's free cash flow turned negative for the first time—at minus 13.8 billion yuan ($2.05 billion). Quarterly cash outflow more than doubled year-over-year, reaching 44.7 billion yuan ($6.65 billion), while total spending amounted to 67.7 billion yuan ($10.07 billion).
The market views such injections with caution, especially against the backdrop of a slowing macroeconomic environment in the second half of the year. Annual AI spending by Chinese tech companies is already equivalent to their cash flow over a year and a half. This is an aggressive bet, but it is necessary for survival in the competitive struggle.
Efficiency as a Trump Card
Notably, the total AI spending of Chinese companies amounts to only one-seventh of that of their American competitors. The gap is explained by limited access to advanced foreign chips and a smaller business scale. However, there is a flip side to this: the cost of training Chinese models is less than 10% of the level of global leaders, and the average API price is less than 20% of international counterparts.
From my expert perspective, in the medium term, the decisive factor will not be the volume of capital investment, but the ability of platforms to effectively monetize their audiences. Chinese companies have already proven that they can achieve goals with minimal resources. Once infrastructure restrictions are lifted, their competitive advantage in efficiency could become a decisive argument in the global AI race. The only question is how quickly the political climate will allow this potential to be realized.