Alibaba’s AI Demand Is Booming. Its Capex Bill Is Now the Bigger Story.

By AltIndex Research · 8 min read · August 20, 2:55 pm

Alibaba’s latest quarter delivered a clean signal that AI demand is becoming real revenue. Its AI cloud and compute revenue rose 45% year over year to RMB 48.44 billion, while capital expenditures climbed 75% to RMB 67.68 billion as it expanded infrastructure for AI workloads.

That combination is the core Alibaba story now. Customers are buying AI capacity, but Alibaba is spending even faster to build it. The investment could establish the company as China’s leading AI-cloud platform. It could also pressure cash flow and returns for years if supply outruns profitable demand.

The Earnings Show Demand. The Capital Spending Shows the Bet.

Alibaba’s June-quarter revenue increased 9% to roughly RMB 269 billion, supported by its core e-commerce operations and faster growth in cloud services. The standout figure was AI cloud and compute revenue, which reached RMB 48.44 billion.

But the investment required to support that growth was enormous. Quarterly capex reached RMB 67.68 billion, or almost $10 billion, reflecting spending on servers, data centers, and compute capacity.

Metric	June quarter	Year-over-year change	What it means
AI cloud and compute revenue	RMB 48.44B	+45%	AI demand is translating into revenue.
Capital expenditures	RMB 67.68B	+75%	Alibaba is building capacity ahead of further demand.
Net income	RMB 10.5B	-75%	The investment cycle is putting pressure on near-term results.

The profit decline should not be read as a one-for-one consequence of capex: infrastructure spending is largely recorded on the cash-flow statement first and flows into earnings over time through depreciation and operating costs. Still, the figures make the economic challenge obvious. Alibaba must turn today’s infrastructure buildout into durable, high-margin cloud revenue before the assets age and competition forces prices down.

Management says it has already deployed roughly half of its previously announced RMB 380 billion, three-year AI and cloud infrastructure commitment. It expects those investments to break even within about three years.

Our Data Shows a More Cautious Picture

The earnings print is strong for the AI narrative, but our broader alternative-data picture remains mixed. Our AI Score for BABA is 39 out of 100, below our 40-point hold threshold.

The main weakness is employment activity. Our job-postings data has declined 35% over the past three months, to 39 open roles, while employee business outlook fell 14.5% with now only 47% of employees reporting that they have a bullish business outlook. Those are not the signals we would normally associate with a company entering a broad-based operating acceleration.

There are meaningful positives. Our estimated web traffic rose 11.2% to 87.5 million monthly visits, and sentiment around the stock is trending moe bullish.

AltIndex signal	Current reading	Three-month change	Read-through
Web traffic	87.5M monthly visits	+11.2%	Customer attention is improving.
Job postings	39	-35.0%	Hiring activity remains the clearest weak signal.
Business outlook	47%	-14.5%	Internal confidence has softened.
Sentiment	80 / 100	+11.1%	Market discussion has become more constructive.
LinkedIn headcount	86K employees	+1.0%	Headcount growth remains a constructive longer-term expansion signal.

The chart below makes that tension visible. BABA’s share price has recovered from its late-June low, but our AI Score remains below the hold threshold. The gap suggests that the market is reacting to the AI-cloud growth story more quickly than our underlying operating signals have improved.

Alibaba’s Own Chips Could Change the Economics

Alibaba’s most important strategic advantage may be its push into self-designed chips. The company is deploying its own silicon at scale to lower the cost of AI compute and reduce dependence on merchant-chip suppliers. If that effort works, Alibaba can improve the return on every yuan it spends on data centers while gaining more control over capacity.

What Investors Should Watch Next

Cloud growth versus capex is the central test. Investors should look for AI cloud and compute revenue to continue outpacing overall company growth, while capex begins to normalize relative to the revenue base it supports.

Cloud margins and free cash flow will show whether the buildout is creating operating leverage or merely adding cost. The company’s three-year payback target gives the market a clear benchmark against which to judge execution.

Custom-chip economics matter because lower compute costs could make Alibaba’s cloud capacity more competitive and more profitable. The company needs to show that its own silicon is improving unit economics, not just expanding supply.

Our employment signals are the alternative-data indicator to watch. A recovery in job postings or employee business outlook would bring the operating picture closer to the improving traffic and sentiment data.

The risk is that Alibaba, Tencent, Baidu, and other Chinese technology companies all build aggressively at the same time. That could create plenty of AI capacity but weaker pricing power. The bull case is equally clear: if Alibaba sustains 40%-plus AI-cloud growth, converts that demand into high utilization, and lowers costs with its own chips, today’s capex could become the foundation of a much larger cloud business.

The Bottom Line

Alibaba has moved beyond AI promises and into AI revenue. The question is no longer whether demand exists. It is whether the revenue can compound quickly enough to justify one of China’s largest AI-infrastructure investment programs.

Sources: Alibaba June Quarter 2026 Results; AltIndex BABA data; reported earnings coverage. Stock prices and our alternative-data readings are as of August 20, 2026.

Disclosure: This article is for informational purposes only and does not constitute investment advice. We aggregate publicly available alternative data signals. Past signal performance does not guarantee future results. Always do your own due diligence before making investment decisions.