[1] Alibaba said it aims to generate more than $100 billion in combined AI and cloud revenue over the next five years, even as the company posted a 67% drop in quarterly profit. The profit decline reflects massive capital reallocation toward AI compute, foundation models (Qwen), and enterprise AI agents—a strategic choice that mirrors Meta's and OpenAI's approach.

[1] In 2026, this manifests clearly: Alibaba is preparing a new enterprise AI offering designed to help companies build and run "agentic" AI systems, expected to be based on Alibaba's Qwen models and linked to the company's broader software and commerce ecosystem. This move shows how China's largest tech groups are trying to turn foundation-model momentum into practical tools for businesses rather than relying only on chatbot demand.

The bigger story is strategic. [2] Enterprise AI is becoming the battleground where cloud, productivity software, payments, workflow automation, and vertical SaaS converge. [2] Alibaba already has reach across commerce, logistics, payments, and workplace software; plugging agentic AI into that stack could give it a powerful distribution advantage in China and potentially in other Asian markets.

Geopolitical context matters here. While OpenAI and Anthropic compete for U.S./Western market share, Alibaba is building an ecosystem optimized for China and emerging markets. The $100 billion revenue target assumes Alibaba can monetize AI across its e-commerce, cloud, and logistics networks—a distribution moat that Western AI companies don't have in China.

My take: Alibaba's bet is different from Western AI companies. They're not trying to build the best frontier model; they're trying to embed AI into every layer of their existing business. The profit drop is temporary pain for long-term optionality. If Qwen becomes competitive with GPT-4 and Claude Opus, and if Alibaba can distribute agentic AI to millions of businesses in Asia, that $100 billion revenue target is credible—not from model API sales, but from monetizing AI-powered workflow automation at the enterprise level.

Sources: