Alibaba’s Plan to Turn AI Investment Into a Full‑Stack Commercial Engine

This growth is helping accelerate cloud performance more broadly, with cloud revenue rising sharply as companies deploy AI workloads on Alibaba’s infrastructure.

The Infrastructure Layer: AI Chips and Cloud Computing

At the foundation of Alibaba’s strategy is computing infrastructure.

The company’s semiconductor work—including chips developed through its chip division—supports AI training and inference workloads across Alibaba Cloud. By integrating AI chips with its cloud infrastructure, Alibaba aims to control more of the AI compute stack, improving cost efficiency and supply resilience for large‑scale model deployment.

This infrastructure layer feeds directly into Alibaba Cloud, which acts as the primary distribution channel for AI services. As enterprises adopt generative AI and agent‑based systems, the resulting demand for compute, storage, and model inference drives increased cloud usage.

The Model Layer: Qwen Foundation Models

Above the infrastructure layer sits Alibaba’s Qwen family of large language models, which the company positions as a core building block for developers and enterprises.

Newer models in the Qwen family are designed for advanced reasoning, coding, and long‑horizon tasks, enabling companies to build complex AI applications and agents directly on Alibaba Cloud.

Alibaba has increasingly positioned Qwen as the centerpiece of its AI ecosystem. The models power consumer AI products while also serving as model‑as‑a‑service infrastructure for developers building AI applications.

The strategy mirrors how major cloud providers package AI capabilities: infrastructure, model APIs, and developer platforms bundled together into a single ecosystem.

The Application Layer: Enterprise AI Agents and Platforms

At the top of the stack are enterprise AI platforms designed to convert AI capabilities into recurring software revenue.

Platforms such as Wukong are being developed to help organizations build and manage AI agents that automate business workflows and decision‑making processes. These systems allow enterprises to deploy multi‑agent AI tools that coordinate tasks across internal systems and data sources.

The idea is to move beyond selling compute and models alone. By enabling companies to build operational AI agents, Alibaba can capture recurring platform and usage revenue tied to enterprise automation.

AI Is Already Driving Cloud Revenue Growth

Alibaba’s financial results suggest that AI adoption is already influencing its cloud business.

• Cloud external revenue growth has accelerated to about 40% year over year.
• AI‑related services account for around 30% of that cloud revenue.
• AI product revenue has grown at triple‑digit rates for ten consecutive quarters.

Independent reporting also shows strong demand for AI‑driven cloud services, with Alibaba’s cloud and AI revenue increasing significantly in recent quarters as enterprises deploy generative AI workloads.

The Next Bet: Enterprise AI Agents

Alibaba believes the next wave of growth will come from AI agents embedded directly into business operations.

Executives argue that AI models are rapidly moving into everyday enterprise workflows, where automated agents can handle tasks ranging from coding and analytics to supply‑chain decisions. Rising token usage and enterprise AI adoption are expected to expand the market for companies that provide end‑to‑end AI platforms.

To capture that opportunity, Alibaba is investing heavily in infrastructure capable of supporting large‑scale AI deployment.

The ¥380 Billion AI Infrastructure Plan

Alibaba has committed to an aggressive infrastructure build‑out, including a three‑year plan to invest about 380 billion yuan (roughly $52 billion) in AI and cloud infrastructure.

The spending is aimed at expanding:

• AI data‑center capacity
• global cloud computing infrastructure
• model training and inference capabilities

Executives have also suggested the company could invest beyond that amount as demand for AI computing continues to grow.

Revenue Targets for the AI Era

The company is betting that this investment will translate into large recurring revenue streams.

Reporting indicates Alibaba expects about 30 billion yuan in AI revenue by 2026, with AI agents projected to drive more than half of future cloud sales as enterprises deploy automation at scale.

While those projections remain forward‑looking, the early indicators—rapid cloud growth and sustained AI product demand—suggest that Alibaba’s AI commercialization push is already underway.

The Big Picture: A Vertically Integrated AI Ecosystem

Alibaba’s approach centers on owning the entire AI value chain:

• AI chips and compute power model training and inference
• Alibaba Cloud distributes infrastructure and AI services
• Qwen foundation models provide the intelligence layer
• Enterprise platforms and agents translate AI capability into business automation

By controlling each layer, Alibaba hopes to capture revenue across the full lifecycle of AI adoption—from compute and model usage to enterprise software built on AI agents.

If enterprise adoption of AI agents accelerates as expected, this vertically integrated strategy could turn what began as heavy AI spending into one of the company’s largest long‑term growth engines.