Alibaba draws Chinese investors on AI, cloud outlook

Mainland Chinese investors have increased their holdings in Alibaba ahead of the company’s Q2 earnings report, according to data from Kaiyuan Securities.

Alibaba saw net share purchases by onshore investors for seven straight sessions via Hong Kong exchange links, raising their stake to 11.1%.

This puts Alibaba’s southbound holdings slightly above Tencent and makes it the most bought stock over the past 60 sessions.

Alibaba shares rose 4.7% on November 24, 2025 after its updated AI tool Qwen reached 10 million downloads in the week after it became available to the public.

Goldman Sachs projects Alibaba’s cloud unit could see 31% year-on-year revenue growth, though overall company revenue is expected to rise by just 1%.

🔗 Source: Bloomberg

🧠 Food for thought

Implications, context, and why it matters.

Alibaba’s GMV share faces pressure from Douyin and Pinduoduo even with AI buzz

By mid-2023, Taobao and Tmall (Alibaba’s core consumer marketplaces) held 44% market share 1. Pinduoduo (a discount, group-buying ecommerce platform) reached 19% 1. Douyin (a TikTok-like short-video app with in-app shopping) logged about $554 billion in Gross Merchandise Value (GMV) in 2024 2.
Goldman Sachs expects Alibaba’s revenue to grow 1% while cloud rises 31% 1. Profit depends on core commerce as shoppers move to social platforms with livestreams and short video shopping 1.
Southbound buying (Mainland investors purchasing Hong Kong-listed shares via Stock Connect) looks tied to AI buzz. Alibaba still faces pressure in its GMV marketplaces from mobile-first, content-driven rivals.

Business demand for Qwen API integrations and fine-tuning is still unclear

Qwen (Alibaba’s large language model family) has 10 million downloads 3. The API offers tool calling, handles multimodal inputs, and supports explicit cache (server-side caching to cut latency plus cost) 3.
Developers plus AI service providers should check Alibaba Cloud’s API pricing tiers, Software Development Kit (SDK) support across languages, and enterprise case studies 4. This will guide demand for integration, compliance, and fine-tuning (customizing a model on proprietary data) services.
Support spans over 100 languages plus specialized models for coding, math, and translation 4. That mix could open global business-to-business (B2B) deals if pricing works for customers.