Alibaba’s net profit dropped 75% last quarter. Free cash flow went deeply negative. And the company’s CEO called it a success. That tells you everything about where Chinese Big Tech is right now.
According to Seoul Economic Daily, Alibaba and Tencent together spent roughly $18 billion on capital expenditure in Q2 2026, with industry estimates putting at least 80% of that figure toward AI infrastructure. Alibaba’s capex alone rose 75% year-over-year to 67.66 billion yuan, driven by expanded compute capacity and rising chip costs. Tencent hit a record 52.784 billion yuan in the same period.
The scale here is hard to overstate. For context, Baidu and Kuaishou, two other well-funded Chinese tech firms, are each spending 6 to 8 billion yuan per quarter on capital expenditure. That’s roughly one-tenth of what Alibaba or Tencent is putting in. The gap between China’s top tier and everyone else is widening fast, and it’s becoming a structural divide, not just a temporary sprint.
Alibaba CEO Wu Yongming made the investment case plainly: there will be a shortage of AI compute through at least 2030, and Alibaba Cloud can recover its infrastructure investment within two and a half to three years. He also flagged that shifting to in-house chips in data centers, rather than commercially procured ones, could improve margins significantly as that transition scales. Alibaba has already committed to spending at least 380 billion yuan on AI and cloud infrastructure over the next three years, so this quarter is one installment in a much larger bet.
Tencent President Martin Lau framed it differently but arrived at the same conclusion. He said the company is comfortable with large AI spending because potential returns are substantial and the downside is protected. His clearest point: Tencent can profit just by leasing out compute capacity, before any AI product revenue kicks in. That’s a useful hedge when product timelines are uncertain.
ByteDance, the third member of China’s top tech tier, is privately held, so its capex isn’t public. But industry estimates suggest it’s spending at a comparable level to Alibaba and Tencent. If that holds, the combined quarterly AI spend from just these three companies likely exceeds $25 billion.
The bigger picture here isn’t just China. This mirrors what Microsoft, Google, and Amazon are doing in the US, where each company is spending tens of billions per quarter on AI infrastructure. What’s different in China is the explicit acknowledgment that near-term profits are being sacrificed deliberately, and that investors are being asked to accept that trade. Alibaba’s 75% profit drop is a direct consequence of that. So far, the market appears to be listening, but the tolerance for sustained losses has limits, and 2027 will be the real test of whether the compute build-out translates into durable revenue.
