Alibaba and Tencent Are Spending Billions on AI — But With Very Different Strategies

This approach treats AI infrastructure as a long‑term strategic asset rather than a near‑term profit driver. Alibaba’s cloud division has become a central pillar of the company’s future growth, with strong expansion in AI‑related cloud services helping drive cloud revenue growth.

However, the strategy has short‑term costs. Heavy spending on AI and cloud infrastructure has weighed on profitability and contributed to weaker earnings results even as the company increases investment.

In practical terms, Alibaba is trying to position itself as China’s primary AI infrastructure provider, similar to how Amazon and Microsoft built dominance through large-scale cloud platforms.

Tencent’s strategy: invest carefully around products and ecosystem

Tencent is pursuing a more measured AI spending strategy.

Instead of prioritising large-scale infrastructure first, Tencent has focused on integrating AI across its existing ecosystem—advertising, gaming, content platforms and enterprise cloud services—while gradually increasing investment in computing capacity.

The company reported Q1 2026 revenue of 196.46 billion yuan, up 9% year‑over‑year, with capital expenditure reaching 31.94 billion yuan, a 16% increase, partly driven by rising demand for AI-related services.

Management has indicated that AI investment will rise further through 2026, especially in the second half of the year, as demand for cloud‑based AI workloads continues to grow.

In contrast to Alibaba’s infrastructure‑first model, Tencent’s strategy ties AI spending more directly to applications and monetisation opportunities across its digital platforms.

Why both companies are increasing capex despite weaker revenue

At first glance, the spending surge may seem counterintuitive. Both companies recently reported revenue results that trailed expectations, and in Alibaba’s case profits have declined as AI investment surged.

Yet management at both firms views AI computing capacity as a strategic constraint rather than a discretionary cost.

Demand for AI services—from model training to inference workloads in cloud platforms—has grown faster than the available supply of GPUs and data‑centre infrastructure. That makes investment in computing capacity a prerequisite for capturing future demand.

In effect, the companies are treating AI infrastructure as a race where falling behind in compute capacity could permanently weaken their position in cloud and AI markets.

The growing role of Chinese‑made AI chips

Another factor accelerating the spending cycle is the shift toward domestic semiconductor alternatives.

Because access to Nvidia’s most advanced AI chips remains uncertain for Chinese companies, firms such as Alibaba and Tencent are increasingly relying on locally developed hardware. These include processors from Huawei as well as chips designed internally by Alibaba’s semiconductor arm.

According to company statements and industry reports, the availability of these domestic chips is gradually increasing, allowing Chinese cloud providers to expand computing clusters and data centres even without unrestricted access to Nvidia GPUs.

The performance of these chips may still lag the most advanced international hardware in some areas—particularly large‑scale model training—but they can significantly expand compute supply for inference workloads, enterprise AI services and cloud platforms.

What it means for China’s AI infrastructure race

The divergence between Alibaba and Tencent highlights two different paths in the AI economy:

• Infrastructure-first strategy: Alibaba is prioritising massive investment in cloud capacity and AI computing power to dominate the infrastructure layer.
• Application-led strategy: Tencent is scaling AI more gradually by embedding it into products across its ecosystem.

Both approaches depend on the same underlying reality: AI growth is constrained by compute supply. As domestic chip production scales and Chinese companies redesign data centres around locally available hardware, the country’s AI infrastructure build‑out could accelerate even without full access to foreign chips.

Whether these investments translate into sustainable profits will depend on how quickly companies can turn AI capacity into commercial services and large‑scale enterprise adoption.