Alibaba group: China's AI Bet Through Cloud Computing Power

Alibaba holds 35.8% of China's AI cloud market, more than ByteDance, Huawei, and Tencent combined. Cloud revenue grew 34% year-over-year in Q3 2025, with AI workloads posting triple-digit growth for nine straight quarters. As China's AI cloud market doubles to $7.3 billion in 2025 and grows 26.8% annually through 2030, Alibaba captures the majority of enterprise spending flowing to infrastructure.

Over 180k AI models built on Qwen now exist on Hugging Face, double the second-place competitor, creating network effects that entrench Alibaba's platform. Qwen's 10 million downloads in week one signal developer adoption at scale. As enterprises deploy AI apps, they choose infrastructure compatible with the most popular models, funneling cloud spending back to Alibaba and making it harder for rivals to gain share.

Alibaba trades at 23.5x 2025 earnings compared to Amazon at 32.1x, despite comparable cloud growth and a dominant position in the world's second-largest economy. The discount reflects regulatory overhang and geopolitical risk, but those fears are priced in while AI infrastructure demand accelerates. If margins expand as cloud mix shifts higher, the valuation gap narrows, offering rerating potential without requiring revenue surprises.

Share Buybacks: Alibaba has $19.1 billion remaining in its buyback authorization through March 2027, after repurchasing $241 million in Q3 2025. If management accelerates the pace to offset dilution and signal confidence in AI investments, it reduces share count and supports valuation as earnings compound.

• Qwen Ecosystem Monetization: With 180k models built on Qwen, Alibaba needs to prove developers using its framework translate into paying cloud customers. As enterprises move from experimentation to production AI deployments, watch for revenue per model to rise and cloud contract sizes to expand, validating the ecosystem lock-in thesis.

• Custom Chip Production: Alibaba's custom chips for smaller AI models reduce US semiconductor dependence and lower inference costs. If chip production scales through 2026 and powers competitive pricing or margin expansion, the company insulates itself from export controls and improves unit economics, strengthening its moat against foreign and domestic rivals.

China AI Adoption Wave: The street forecasts China's AI adoption will exceed 30% by 2030, with AI-related spending reaching 1% of annual GDP. As traditional industries digitize and the government pushes AI integration, Alibaba's infrastructure position captures enterprise spending at scale, creating a structural tailwind that compounds for years.

US export controls block Alibaba from accessing advanced AI chips, forcing reliance on custom alternatives and smuggled inventory. While the company trains competitive models, chip shortages could constrain deployment at scale, raising inference costs and slowing enterprise adoption. If restrictions tighten further or enforcement improves, Alibaba's ability to meet surging AI demand weakens, handing market share to competitors with better chip access.

Beijing fined Alibaba $2.78 billion in 2021 for antitrust violations and maintains oversight despite completing a rectification process in 2024. The Pentagon recently suggested adding Alibaba to a list of companies aiding China's military, raising delisting risk and capital flight concerns. Regulatory shifts remain unpredictable in China's civil law system, where penalties lack clear precedent and political priorities can override commercial logic, creating valuation overhang.

ByteDance's Volcano Engine holds 14.8% of China's AI cloud market, Huawei Cloud captures 13.1%, and Tencent Cloud takes 7%, all investing heavily to close the gap. ByteDance's Doubao model gained 14.1% enterprise adoption versus Alibaba's Qwen at 17.7%, showing rivals can narrow leads quickly. If Alibaba's $52 billion spending fails to widen its moat, pricing pressure rises and margin expansion stalls, undermining the cloud growth thesis.