Wei Xiong, an analyst specializing in China's Internet sector at UBS Securities, remarked that, when compared to the United States, China faces a lower risk of an AI bubble forming. Currently, there are no clear indicators of such a bubble. From a financing standpoint, leading Chinese AI model firms primarily depend on financial support from their parent companies for research and development (R&D). Incidents of circular financing and mutual dependencies are less common, which guarantees a steady stream of funding. Regarding capital investments, major Chinese Internet enterprises exhibit a more pragmatic and cautious stance, prioritizing return on investment (ROI) and the efficiency of computing resource utilization. By 2025, the capital expenditures of leading Chinese firms are projected to be just one-tenth of those of their U.S. peers, yet they are expected to attain similar model performance levels. Furthermore, these companies adopt a measured and incremental strategy in constructing their own data centers. Since the latter half of 2024, data centers have maintained consistently high utilization rates, driven by authentic AI-related workloads.