Key takeaways

  • China's Alibaba reported a 75 per cent fall in quarterly net profit on Thursday as the tech giant heavily ramped up AI capital expenditure…
  • Demand for the cloud computing power needed to train and run enterprise AI systems has surged, benefiting China's largest technology…
  • 6 per cent in early morning trading.

What happened

China's Alibaba reported a 75 per cent fall in quarterly net profit on Thursday as the tech giant heavily ramped up AI capital expenditure, betting future growth on its enterprise cloud and AI model services. The group reported a 9 per cent rise in revenue for April-June, as strong AI demand fuelled growth ​in its cloud business.

Domestic AI, chip ​raceAlibaba is locked in a battle with other Chinese tech giants and startups to release more capable, low-cost frontier AI models, highlighting the rapid pace of advancement of Chinese AI models and their shorter release cycles.

Alibaba's fintech affiliate Ant Group reported 1 per cent year-on-year growth in quarterly profit, Reuters calculations showed, as it has attempted to pivot towards agentic AI commerce, AI digital health applications and embodied AI models in recent years.

Why it matters

Demand for the cloud computing power needed to train and run enterprise AI systems has surged, benefiting China's largest technology companies. 4 ‌billion) AI investment planned ⁠for 2026-29. "As we ramp up deployment ​of our own proprietary chips in our data centres ... 44 billion yuan in the June quarter, backed by strong growth in its AI ​model-as-a-service business, which has surpassed 16 billion yuan in annual recurring revenue.

6 per cent in early morning trading. 68 billion yuan ​in April-June, the group's first quarter, as it increased procurement of CPU chips due to AI agent demand and ⁠semiconductor component prices went up.

What to watch

Advanced chips produced by Alibaba's in-house unit T-head are already being deployed at scale on "supernodes" - massive server racks linking hundreds of semiconductors together - for AI model training and ‌inference, Wu said. The company is banking on scaling up deployment of its in-house chips to reduce capex costs and to generate future revenue.