Alibaba Earnings: AI Cloud Jumps 45%, Profit Drops 75%

Alibaba reported the quarter it has been telling investors to expect: cloud is accelerating on AI demand, and the bill for that growth is landing on the bottom line. After results on Wednesday, August 20, 2026, the Chinese e-commerce and cloud group posted its fastest cloud growth in years alongside a 75% drop in net income, as capital spending on AI infrastructure climbed by three-quarters year over year. US-listed shares slipped in the reaction, extending a pattern that has defined this earnings season: the market is scrutinizing what AI growth costs, not just celebrating that it exists.

The quarter by the numbers

For the three months ended June 30 — the first quarter of Alibaba’s fiscal 2027 — revenue rose 9% year over year to RMB 268.95 billion, or about $39.64 billion. That top-line figure was solid but not the story. The story was the split between an accelerating cloud franchise and a sharply lower profit.

Net income fell to RMB 10,444 million (roughly $1.54 billion), a decline of 75% from a year earlier. Alibaba attributed the drop to a decrease in income from operations, lower net gains from the disposal of investments, and a decline in mark-to-market gains on its equity holdings — a reminder that a large share of the year-ago profit came from investment marks rather than the core business. On an adjusted basis, earnings of 8.52 yuan per American depositary share missed analyst estimates of about 10.53 yuan, and US-listed shares fell roughly 2.6% in premarket trading, with declines of around 5% at points during the session.

The offsetting bright spot was the cloud unit. Alibaba Cloud’s external revenue growth accelerated to 45%, its fastest pace in 22 quarters, driven by adoption of the company’s AI products. The AI Cloud and Compute Services segment posted revenue of about RMB 48.4 billion ($7.14 billion) for the quarter.

AI revenue and the Qwen engine

The metric Alibaba most wants investors to anchor on is its AI product revenue, which maintained triple-digit year-over-year growth for the 12th consecutive quarter, reaching about RMB 12.4 billion ($1.82 billion). Chief Executive Eddie Wu said the annualized revenue run rate from AI-related products has surpassed RMB 49.5 billion (about $7.34 billion), and that AI’s share of Alibaba Cloud’s external revenue rose to 35% during the quarter — up sharply as enterprises route inference and training workloads through the platform.

Underpinning that demand is the Qwen model family. Earlier in August, Alibaba unveiled Qwen3.8-Max, which it described as its most powerful model to date, and the broader Qwen line-up has become one of the most widely deployed open-weight families in the industry. The commercial read-through is direct: the more capable and widely used the models, the more compute customers rent — and the more revenue flows through the cloud segment that Alibaba is now positioning as its primary growth engine.

That flywheel is not free. The company disclosed that its AI Labs and Applications segment posted an adjusted EBITA loss of about RMB 13.86 billion ($2.04 billion) for the quarter, widening from a RMB 3.22 billion loss a year earlier. The larger loss reflects heavier investment in model development and rising inference costs tied to the consumer-facing Qwen app — the same kind of front-loaded cost that hyperscalers everywhere are absorbing to seed usage.

Capex climbs, and a RMB 380 billion commitment

The clearest signal of Alibaba’s intent is its spending. Capital expenditures reached RMB 67.68 billion (about $9.98 billion) in the quarter, a 75% increase from the same period a year earlier, as the company poured money into data centers, servers, and networking for AI. That single-quarter figure puts Alibaba on a spending trajectory that would have been unthinkable for a Chinese internet company two years ago.

It also advances a commitment the company made in February 2026 to invest RMB 380 billion — roughly $53 billion — over three years to build out cloud and AI hardware infrastructure. The June quarter shows that plan moving from press release to purchase orders. Like its US counterparts, Alibaba is building capacity ahead of a demand curve it believes is still steepening, accepting near-term margin compression in exchange for a position in what it treats as a winner-take-most race.

The dynamic mirrors what has played out among the American hyperscalers this year. When Alphabet raised its capex forecast alongside 82% cloud growth, its stock fell despite the beat; the market has repeatedly rewarded AI revenue and then sold the shares on the cost of producing it. Alibaba’s print fits the template — accelerating cloud, a heavy and rising infrastructure bill, and a share-price reaction that focused on the second half of that sentence.

A different regulatory and competitive backdrop

Alibaba’s AI push runs on a distinctly Chinese set of constraints and advantages. Export controls limit access to the most advanced US accelerators, pushing Chinese platforms toward domestic silicon and toward squeezing more out of the chips they can obtain — a pressure that has, in turn, made efficient open models like Qwen strategically central rather than a side project. On the demand side, Chinese enterprises and developers increasingly standardize on domestic models for compliance and cost reasons, and Alibaba Cloud is the largest beneficiary of that shift.

The Qwen family’s reach now extends beyond China’s borders. Earlier this year, Apple’s plan to use Qwen for its China-market AI features underscored how a domestic model can become critical infrastructure for foreign firms operating in the country. Each such deployment reinforces the same loop that showed up in this quarter’s numbers: more model usage, more cloud consumption, more justification for the RMB 380 billion buildout.

What it means

Alibaba delivered a quarter that looked, on the fundamentals, like an inflection — cloud growth at a 22-quarter high, AI revenue compounding in triple digits for three straight years, and a run rate that is starting to matter to the overall business. The stock softened anyway, because the same quarter carried a 75% profit drop and a 75% jump in capex. That gap between the operating story and the reported profit is the whole picture.

Who wins. The cloud and AI franchise is the clear winner internally, and Alibaba is reallocating the company’s resources toward it in real time. Domestic chip suppliers, data-center partners, and power providers gain a customer that just committed to spending at hyperscaler scale. And if China’s enterprise AI adoption curve holds, the capacity going in now becomes a durable moat that smaller domestic rivals cannot match — the same logic driving the global AI capex boom.

Who feels it. Shareholders underwriting near-term earnings take the hit. Much of the year-ago profit came from investment gains that did not repeat, and the core business is now absorbing a widening AI-applications loss plus a heavier depreciation load as new capacity comes online. A company that once threw off large, predictable profits is choosing, deliberately, to look less profitable while it builds.

What to watch next. Three things. First, whether cloud growth holds near the mid-40s — the 45% external-revenue pace is what justifies the RMB 380 billion plan, and any deceleration would reframe the spending as overreach. Second, the AI Labs loss trajectory, since that line shows whether consumer AI is a strategic seed or an open-ended subsidy. Third, the capex-to-cash relationship over the next few quarters, because as the economics of AI data centers make plain, this is spending that is front-loaded with cost and back-loaded with return — and investors are deciding, quarter by quarter, whether Alibaba’s bet is a moat or a money pit.