Alibaba has crossed a critical threshold: its artificial intelligence business is now profitable and growing fast, with AI-related revenue hitting 9 billion yuan in early 2026. Yet despite this milestone, investors remain skeptical. The company's stock has fallen 11.31% this year, while competitors like Anthropic have soared to a $900 billion valuation. The gap between Alibaba's growth and its profit margins reveals why the capital markets are treating China's AI ambitions differently than their American counterparts.

The shift toward profitable AI is real across the industry. Leading U.S. companies have already reached the inflection point where AI generates serious money. Anthropic, which sells access to its Claude AI model and AI agents for enterprise tasks, is expected to reach $10.9 billion in revenue for the second quarter of 2026, a 130% year-over-year increase. More importantly, the company achieved an operating profit of $559 million for the first time, with computing costs dropping to just 56 cents per dollar earned.

Google has taken a different path, leveraging its full-stack advantage. The company sells computing power, AI tokens, and applications through Google Cloud, which generated $20 billion in revenue during the first quarter of 2026, up 63% year-over-year. Google Cloud's operating profit soared to $6.6 billion, and the company reported a backlog of cloud business orders exceeding $460 billion, nearly double the previous quarter.

Why Is Alibaba's Profit Margin So Much Lower Than Google's?

Alibaba is pursuing a similar strategy to Google, positioning itself as a full-stack AI company with chips, cloud infrastructure, and models all working together. The company's self-developed Zhenwu M890 AI chip delivers three times the performance of its predecessor, the Zhenwu 810E, and is deeply integrated with Alibaba's Qwen series of large language models. The newly released Qwen-3.7-Max has strengthened programming capabilities, long-task processing, and general intelligent agent abilities.

Yet the numbers tell a different story. In the first quarter of 2026, Alibaba Cloud's adjusted profit margin was just 9.1%, compared to Google Cloud's 33%. This gap exists for concrete reasons. Alibaba's foundational AI models score 1,475 points on comprehensive benchmarks, trailing Anthropic's Claude-Opus-4.6-Thinking at 1,502 points and Google's Gemini-3.1-Pro-Preview at 1,488 points. Weaker base models mean less efficient AI agents and lower customer satisfaction.

Beyond Alibaba's own limitations, the entire Chinese AI industry faces structural challenges. Domestic AI chips still lag behind Nvidia's GPUs and Google's TPUs in comprehensive performance, software ecosystem maturity, and cluster interconnection capabilities. These gaps make it harder for Chinese companies to deliver the same level of efficiency and cost-effectiveness that American competitors achieve.

How to Understand Alibaba's Path to Higher Profitability

• Model Capability Improvements: Alibaba is investing heavily in strengthening the Qwen series of large language models to match or exceed competitor benchmarks, which would enable more efficient AI agents and higher customer satisfaction.
• Chip Manufacturing Scale: Pingtouge, Alibaba's chip subsidiary, has achieved large-scale mass production of self-developed GPU chips, reducing dependency on external suppliers and improving margins over time.
• MaaS Platform Growth: Alibaba's Bailian MaaS (Model-as-a-Service) platform, which allows enterprises to access and customize AI models, is expected to generate over 10 billion yuan in annual recurring revenue by the second quarter of 2026 and exceed 30 billion yuan by year-end.

During Alibaba's earnings call, Wu Yongming, a company executive, released optimistic projections about the path forward. He noted that demand for API tokens (the units used to access AI models) is essentially unlimited, and the MaaS business currently has higher gross profit margins while facing supply constraints. "The gross profit margin of Alibaba Cloud will increase significantly in the next 1 to 2 years," Wu stated, signaling confidence that the company can narrow the gap with Google.

"The gross profit margin of Alibaba Cloud will increase significantly in the next 1 to 2 years," said Wu Yongming.

Wu Yongming, Alibaba Executive

Alibaba is not alone in pursuing this strategy. Volcengine, ByteDance's cloud division, set a revenue target exceeding 10 billion yuan for its MaaS business in 2026, showing that Chinese AI companies are converging on the same enterprise-focused model.

What Makes Enterprise AI So Profitable?

The shift toward business-to-business AI services reflects a fundamental reality: enterprises are desperate for AI solutions that reduce costs and improve efficiency, and they are willing to pay premium prices for tools that deliver measurable returns. In financial and tax scenarios, AI agents perform at 10 times the efficiency of traditional accountants while reducing labor costs by 60%. In programming, Anthropic's engineers used Claude Code to complete the transcription of 50,000 lines of Scala code in just 4 days, a task that would have taken at least 70 days using traditional methods.

Anthropic has built its entire business around this insight. Since its founding, the company has focused precisely on AI programming and enterprise users in high-value industries such as finance, technology, and professional services. The strategy is working: 34.4% of U.S. enterprises have paid for Anthropic's services, exceeding OpenAI's adoption rate of 32.3%. More than 1,000 enterprise customers now spend over $1 million annually with Anthropic, a figure that doubled in just three months.

Alibaba's Bailian MaaS platform ranks first in China by revenue scale, and the company has seized a first-mover advantage in enterprise applications through DingTalk, its workplace collaboration platform with over 800 million registered users and 26 million enterprise organizations. The company recently launched Wukong, an enterprise office intelligent agent fully integrated with DingTalk, extending AI capabilities directly into the workflows where businesses operate daily.

The question now is whether Alibaba can execute fast enough to close the margin gap. If the company succeeds in strengthening its foundational models, scaling chip production, and capturing more enterprise customers through DingTalk and Bailian, its stock price could see significant upside. But investors are waiting for proof that the company can match the efficiency and profitability of Google and Anthropic, not just match their growth rates.

" }