Alibaba’s AI Bet Reaches Its Payoff Test as Cloud Growth Hits 45%

That outlook remains a forecast, not a guarantee. Slower enterprise adoption, weaker pricing or excess computing capacity could reduce the revenue generated by each yuan invested.

Higher utilisation is starting to support margins

AI-cloud adjusted EBITA margin increased to 11.6%, compared with roughly 7% a year earlier. The improvement matters because data centres, chips and other computing infrastructure create substantial fixed costs. As more customers use that capacity, revenue can grow faster than the cost of operating the existing platform.

Further margin expansion would make the investment case stronger in two ways. It would increase the profit contribution from each additional unit of AI revenue, and it would shorten the time required to recover the cost of the computing infrastructure. Alibaba’s return on spending therefore depends not only on revenue growth, but also on the quality of that revenue and the utilisation of its AI fleet.

The two- to three-year payback case is conditional

CEO Eddie Wu said Alibaba expects its AI-related capital expenditure to break even within three years under current average gross margins, with a possible two-year outcome if gross margins improve further. The mechanism is straightforward: higher utilisation, improving gross margins and a larger mix of AI models, applications and other higher-value services would allow the investment to generate enough contribution to cover its cost.

This should be read as a management outlook rather than a realised return. Alibaba has reached the halfway mark of the planned investment while still carrying the financial burden of the build-out. The company may need continued strong growth before the economics of the infrastructure become visible in group-level earnings.

Why the headline profit decline still matters

The cloud trajectory does not eliminate the cost of funding the expansion. Reuters reported that Alibaba had spent half of its 380 billion yuan AI investment plan for 2026–29, while quarterly capital expenditure increased 75% and net profit declined 75% year on year. AP reported quarterly profit of 10.5 billion yuan, down from 43.1 billion yuan a year earlier, even as AI-related revenue rose 45%.

That contrast creates the central investor question: can future cloud gross profit grow quickly enough to offset the cash and earnings sacrificed today? A strong answer would require several conditions to hold at once:

• AI and cloud demand must remain strong enough to keep new capacity well utilised.
• Gross margins must continue improving as the business scales.
• Pricing competition must not erode the return on computing capacity.
• Capital spending must eventually grow more slowly than AI-related revenue and profit.

What Alibaba’s halfway point means

Alibaba’s progress supports a cautiously positive interpretation of its AI spending. Cloud growth is accelerating, AI-cloud margins are expanding and management sees a path to break-even in roughly two to three years.

But the company is still in the investment-heavy part of the cycle. For now, the clearest result is faster cloud growth—not a completed return on the full infrastructure commitment. The next decisive indicators will be whether growth moves above 50% as forecast, whether margins continue to widen and whether Alibaba can reduce the gap between rising AI revenue and the cash cost of building the capacity behind it.