Alibaba profit falls 75% after ramping up AI infrastructure spending

China's Alibaba reported a 75% decline in quarterly net profit in June due to heavy investments in AI infrastructure, as the company bets on future growth from its enterprise cloud and AI model services. Despite a 9% revenue increase for the April-June quarter, strong demand for AI cloud computing power boosted Alibaba's cloud business.

CEO Eddie Wu revealed plans to break even on AI-related capital expenditures within three years based on current average gross margins, after spending half of the company's 380 billion yuan ($56.4 billion) AI investment for 2026-29. To secure future growth, Alibaba intends to increase AI infrastructure investments and replace commercially procured chips with proprietary ones, anticipating higher gross margins and profitability.

The enterprise AI cloud and compute services revenue rose 45% to 48.44 billion yuan, driven by the AI model-as-a-service business, which surpassed 16 billion yuan in annual recurring revenue. However, adjusted earnings per American Depositary Share fell short of estimates at 8.52 yuan, and U.S.-listed shares dropped 4.6% in early trading.

Alibaba's capital expenditure rose 75% to 67.68 billion yuan in the first quarter, driven by increased CPU chip procurement due to demand for AI agents and rising semiconductor component prices. The company reported revenue of 268.95 billion yuan in the first quarter, aligning with analysts' estimates. As competition intensifies among Chinese tech giants and startups to develop more capable, low-cost AI models, Alibaba is focusing on offering better agentic and coding capabilities for the Chinese domestic enterprise market.

The company is also a significant investor in Chinese frontier AI startups and supplies cloud computing infrastructure to Moonshot. Alibaba's in-house unit T-head is deploying advanced chips at scale for AI model training and inference on supernodes. Despite current monetization challenges, Alibaba remains committed to achieving artificial general intelligence (AGI) and expects to reach that goal by investing in developing frontier AI models.

Alibaba recently reorganized its businesses into four main units and tasked Wu with leading the Alibaba Token Hub group to make its AI segment profitable.