NextFin News - Alibaba Group's net income collapsed 75% in the quarter ended June 30, 2026, to RMB10.44 billion (US$1.54 billion), as China's e-commerce and cloud giant accelerated spending on artificial-intelligence infrastructure at a pace that is now consuming the profits of its core online-shopping business. Revenue rose 9% to RMB268.95 billion (US$39.64 billion), but capital expenditures jumped 75% to RMB67.68 billion (US$9.98 billion), and free cash flow swung to a RMB44.67 billion (US$6.58 billion) outflow - more than double the prior-year drain. The report, released August 20, 2026, frames the question now facing investors: is Alibaba's AI bet a cyclical squeeze before the payoff, or a permanent shift into a capital-heavy model where the old profit engine can no longer keep up?
Alibaba's U.S.-listed shares fell after the release, unwinding part of a rally that had lifted the stock roughly 4% earlier in the month on enthusiasm for the company's latest AI model launch. The market had priced in a modest earnings dip; what it got was a profit plunge paired with a spending cycle that shows no sign of slowing.
The Numbers: Revenue Grows, Profits Evaporate
The headline figures from the earnings release draw a stark picture. Revenue of RMB268.95 billion (US$39.64 billion) grew 9% year over year, ahead of the roughly $38.6 billion revenue consensus tracked ahead of the print. But profitability measures deteriorated across the board: income from operations fell 57% to RMB15.16 billion (US$2.23 billion); adjusted EBITA, the non-GAAP profit measure investors watch, dropped 30% to RMB27.33 billion (US$4.03 billion); and non-GAAP net income declined 38% to RMB20.72 billion (US$3.05 billion). Non-GAAP diluted earnings per American depositary share fell 42% to RMB8.52 (US$1.26).
The divergence between the top line and the bottom line is the story. Revenue grew because AI demand is real and accelerating. Profits fell because meeting that demand requires spending money today for revenue that arrives later - and in much larger quantities than the market had modeled.
Capital expenditures reached RMB67.68 billion (US$9.98 billion), up 75% from the year-ago quarter, driven by spending on AI infrastructure. Free cash flow - operating cash flow minus capital spending - was a negative RMB44.67 billion (US$6.58 billion), compared with a negative RMB18.82 billion in the same quarter of 2025. Operating cash flow itself rose 11% to RMB22.95 billion (US$3.38 billion), which means the cash-generation engine of the business still works; it is simply being outspent.
Management framed the quarter as a deliberate trade-off. Chief executive Eddie Wu said in the release:
We delivered a strong quarter, driven by the improving commercialization of our full-stack AI capabilities. Alibaba Cloud's external revenue growth accelerated to 45%, with AI-related product revenue delivering triple-digit growth for the twelfth consecutive quarter.
Chief financial officer Toby Xu added:
This quarter, we delivered robust revenue growth and margin improvement across our core businesses. Cloud segment revenue growth continued to accelerate, with quality earnings and operating leverage increasing EBITA margin to 12%.
The AI Bill Is Arriving Faster Than the Revenue
The mechanism behind the profit collapse is straightforward, and it is the same mechanism every hyperscaler faces when an AI buildout begins: the cash goes out years before the revenue comes in at scale. Training frontier models requires clusters of advanced processors; serving those models requires data-center capacity; and both must be purchased before customers sign contracts. In cloud computing, you build the capacity and then rent it. In AI, you must keep rebuilding it, because the models themselves depreciate on a timeline measured in months, not years.
Alibaba's spending is concentrated in the AI Labs and Applications segment, which houses its Qwen model development, the Qwen consumer app, and the QwenWork enterprise agent launched this quarter. That segment posted an adjusted EBITA loss of RMB13.86 billion (US$2.04 billion), compared with a loss of RMB3.22 billion a year earlier - a more than fourfold widening in twelve months. Meanwhile, AI cloud and compute services revenue rose 45% to RMB48.44 billion. The ratio tells the tale: for every yuan of AI cloud revenue, the company is spending multiples on the infrastructure and model development that will (or will not) generate the next yuan.
Here is the hook that the headline number hides: a single quarter's capital expenditure of RMB67.68 billion annualizes to roughly RMB270 billion - more than twice the implied annual pace of the RMB380 billion, three-year AI and cloud commitment the company announced in February. In other words, the commitment that sounded enormous six months ago is already tracking as a floor, not a ceiling. At a developer conference in August, Wu indicated that demand for AI infrastructure and the speed of industry development had far exceeded the company's own anticipation, and that the existing investment plan may prove too small.
The Core E-Commerce Engine Is No Longer a Bottomless ATM
The second half of the story is what makes this profit dive harder to absorb than a normal investment cycle. For most of Alibaba's history, the company's structure was simple: the China e-commerce business printed cash, and that cash funded everything else - logistics, international expansion, entertainment, and now AI. The investment cycle was self-financing.
That cushion is thinner now. Customer management revenue - the advertising and commission income from merchants on Taobao and Tmall, the business that has always been the profit center of the group - fell 7% year over year. Excluding the contra-revenue impact of a new business-development program, it grew just 1% on a like-for-like basis. A core cash cow that grows at 1% cannot subsidize a capital cycle that grows at 75%.
And e-commerce is not merely stagnating; it is itself under investment pressure. The company has committed RMB50 billion to its quick-commerce push, and chief financial officer Xu Hong told analysts on the earnings call that this spending is incremental to ongoing investments in users and supply chains. So Alibaba is running two investment cycles at once: AI on one side, consumption on the other. Xu framed the quick-commerce logic as focused on "the overall incremental benefits to our e-commerce business," arguing that the delivery war drives user growth and higher take rates. That may be true. It is also, in the near term, another claim on the same pool of profits.
This is the structural shift beneath the cyclical squeeze. The profit pool that once made Alibaba's diversification affordable is itself under pressure from a weak Chinese consumer and intensifying competition. The company is not choosing between investing and returning cash because it wants to; it is investing because not investing in AI would be worse, while the business that used to pay for everything is no longer growing fast enough to make the choice comfortable.
Cyclical Squeeze or Structural Regime Change?
Getting this call right determines the entire investment thesis, so it is worth separating the two forces rather than blending them.
The cyclical leg is real and should not be dismissed. Capital expenditure in infrastructure is front-loaded by definition: you buy the chips and build the data centers before the customers arrive. Cloud revenue growth is accelerating, not decelerating - external revenue up 45%, AI-related products in triple-digit growth for twelve consecutive quarters, cloud EBITA margin holding at 12%. History offers a template: Amazon Web Services and Microsoft Azure both absorbed years of heavy infrastructure spending before their margins inflected upward. If Alibaba's AI monetization follows that arc, today's profit dive is the price of admission to a larger business, and margins recover as utilization catches up with capacity.
But the structural leg is equally real, and it is the one the market is not pricing. AI is not simply cloud 2.0. In classic cloud, capacity is durable - a server rack built today serves customers for years. In AI, the capacity is the model, and the model depreciates on a research timeline. Qwen 3.8 Max launched this year as the company's flagship; the next flagship is already in training. The competitive bar is set by Tencent's WeChat agent and ByteDance's Doubao, China's most popular AI app, both of which are pushing subscription and agentic models. To stay in the race, Alibaba cannot build once and rent forever; it must spend continuously, and the spending scales with usage because inference costs rise with every query.
My judgment: both forces are at work, and they point in opposite directions across time horizons. The near-term leg is cyclical - front-loaded capex, temporary margin compression, a revenue curve that is still accelerating. The long-term leg is structural - Alibaba is transitioning from an asset-light platform company, which commanded a premium multiple, into an asset-heavy AI utility, which will be judged on return on invested capital. That re-rating is the hidden cost of the AI pivot, and it is not captured in any single quarter's earnings.
What the Market Priced In - and What It Hasn't
Consensus expectations going into the print called for revenue of roughly $38.6 billion and a mid-single-digit earnings decline. Alibaba delivered revenue of $39.64 billion, a slight beat, but a 75% collapse in GAAP net income and a 42% decline in non-GAAP earnings per share. The market priced a modest earnings dip, not a profit halving.
But the market also knew capex was rising - the RMB380 billion commitment was public. The second-order question is not whether this quarter missed; it is what a sustained free-cash-flow outflow does to the shareholder-return engine that supported the stock through its 2025 rally. Alibaba bought back shares aggressively in prior quarters, and buybacks were a meaningful contributor to per-share growth. A free-cash-flow outflow of RMB44.67 billion in one quarter, if repeated, limits that capacity. It also makes the stock more sensitive to the China risk premium and less sensitive to the AI narrative, because investors will discount cash flows that may never arrive rather than reward a story that has not yet paid.
The cross-asset transmission runs further. If Alibaba's free cash flow stays negative, the company's enormous cash pile - RMB474.51 billion (US$69.93 billion) as of June 30 - becomes a buffer rather than dry powder for acquisitions or aggressive buybacks. That shifts the investment case from a total-return story (growth plus capital returns) to a pure option on AI monetization. Options are volatile. The stock's roughly 4% rally earlier in August on the Qwen launch, unwound on the earnings print, is a preview of that volatility.
The Strongest Case Against This View
The bear case is not a strawman, and it deserves its full weight. It rests on four observable facts from this very release: free cash flow is deeply negative at RMB44.67 billion in a single quarter; the core e-commerce cash cow is growing at roughly 1% on a like-for-like basis; the AI unit's adjusted EBITA loss widened more than fourfold to RMB13.86 billion; and the company is simultaneously funding a RMB50 billion quick-commerce investment cycle. A mainstream read of these facts - one shared by analysts who have flagged the tension between Alibaba's AI ambition and its deteriorating cash returns - is that the company is caught between two capital-intensive battles while its profit engine stalls, and that the RMB380 billion commitment is merely the opening tranche of a much longer spend. If that read is right, the current valuation, which still embeds a recovery in earnings power, is resting on a cyclical story that is actually structural.
The answer to the bears is balance-sheet depth, not earnings power. With RMB474.51 billion in cash and liquid investments, Alibaba can fund this cycle for years without solvency risk. The question is not whether the company can afford to play; it is whether the game pays. And here the bull case has a real answer: cloud revenue growth is accelerating, AI-related revenue is growing in triple digits, and management expects AI to exceed 50% of cloud revenue within a year, up from roughly 30% in the March quarter. If that monetization arrives before the balance sheet weakens, the current profit dive is a trough, not a destination.
So the falsifying signal is specific and observable: if free cash flow remains an outflow greater than RMB40 billion per quarter for the next two consecutive quarters while AI-related revenue growth decelerates below triple-digit year-over-year growth, the "temporary squeeze" thesis is wrong. At that point, the spending is not front-loaded investment; it is a structural cost of doing business in AI, and the valuation must be judged accordingly. Until then, the burden of proof sits with the bears to show that the revenue acceleration is not real.
Outlook: Three Horizons, Three Scenarios
Short term (sentiment and liquidity): negative. The 75% profit dive and the negative free-cash-flow print will weigh on the stock, and the early-August rally built on AI enthusiasm has already been reversed. Expect volatility to remain elevated into the next print.
Medium term (fundamentals): watch two metrics. First, whether AI-related revenue exceeds 50% of cloud external revenue within a year, as management guides. Second, whether cloud EBITA margin holds near the 12% level even as AI mix rises - evidence that the company is monetizing AI at attractive margins, not just buying revenue with subsidized compute.
Long term (structural): Alibaba is becoming an AI utility. That is a different company from the asset-light platform that listed in 2014, and it will be valued on different terms - return on invested capital, capex intensity, and the durability of its model advantage rather than just user growth and take rates.
Three scenarios frame the path. The base case: revenue grows in the high single digits, capex moderates toward RMB120-130 billion per year, free cash flow turns positive by late fiscal 2027, and the stock trades on a sum-of-the-parts discount to peers. The upside case: AI monetization accelerates faster than capex, cloud becomes majority-AI revenue ahead of schedule, operating leverage kicks in, and the multiple re-rates toward global cloud peers. The downside case: capex stays above RMB150 billion per year, Chinese consumption stays weak, AI competition forces more spending with lower returns, and free cash flow remains negative through fiscal 2028.
What to watch next: the capex and free-cash-flow figures in the September-quarter report; the trajectory of customer management revenue as the tell on core e-commerce health; and any revision to the RMB380 billion investment commitment, which would signal whether management sees the spending cycle peaking or still climbing.
Alibaba is not losing the AI race; it is paying the entry fee in real time. The 75% profit dive is the bill for a seat at the table - and the question is no longer whether Alibaba can afford to play, but whether the game it is buying into will ever pay like the e-commerce business it is leaving behind.
Data as of the Alibaba Group earnings release for the quarter ended June 30, 2026, published August 20, 2026. Market levels as of the August 20 trading session.
Explore more exclusive insights at nextfin.ai.
