Alibaba's AI Push Hits a Wall: Disclosure Problem
Alibaba reported strong cloud revenue growth and improving margins in its latest earnings, but investors face a critical problem: they cannot determine how profitable the company's artificial intelligence business actually is.
The Chinese technology giant's cloud and AI division surged 38 percent year-on-year to roughly $6 billion in quarterly revenue. Adjusted profit for that unit jumped more than 50 percent. Yet the company still reported an operating loss, partly because massive AI infrastructure spending is hidden inside a broad accounting category called "All Others."
That segment groups unrelated operations-logistics, mapping, video games, supermarkets, and AI model training-making it impossible for investors to isolate how much Alibaba actually spends on artificial intelligence development. The "All Others" segment recorded adjusted losses exceeding 500 percent year-on-year.
Why This Matters for Your Work
As an IT or development professional, this disclosure problem reveals something important: companies racing to build AI infrastructure are spending billions on costs that remain largely invisible to investors. Understanding where those dollars go-and whether they generate returns-is becoming essential for anyone evaluating AI investments or infrastructure decisions.
Alibaba is betting heavily on cloud computing and AI as its next major growth engine. Chief Executive Eddie Wu has set a target of generating around $100 billion in annual AI revenue within five years. The company is investing across cloud infrastructure, model training, the Qwen large language model platform, and enterprise AI services.
The Real Cost of Building AI Models
Training advanced AI models is extremely expensive. Large language models require enormous computing power, specialized semiconductors, large-scale data center infrastructure, and massive electricity consumption. Research personnel, algorithm development, model testing, and AI safety systems add further costs.
Global AI leaders-OpenAI, Google, Microsoft, Meta, and Anthropic-all spend billions annually on AI infrastructure and model development. Alibaba now faces similar pressures as it attempts to build competitive AI systems within China.
Alibaba's Qwen platform has become one of China's most important domestic AI systems. The company is using it to power chatbots, enterprise AI tools, productivity software, developer platforms, and coding assistants. Qwen also represents part of China's broader effort to develop local AI alternatives to Western systems like ChatGPT and Google Gemini.
Why Cloud Infrastructure Matters Now
Cloud computing has become central to Alibaba's strategy. Historically, the company generated most revenue from e-commerce inside China. But cloud computing and AI services are expected to account for a much larger share of revenue over time.
AI systems require enormous computing infrastructure. Companies developing AI applications need access to data centers, GPU processing power, AI-optimized servers, storage systems, and scalable computing resources. Alibaba is positioning itself as both an AI developer and a foundational infrastructure provider for the Chinese AI ecosystem.
This dual role matters. If Alibaba succeeds, it could eventually generate more revenue from cloud and AI services than from its original e-commerce business.
The Competition Problem
Alibaba faces intense competition inside China. Tencent, Huawei, Baidu, and ByteDance are all investing aggressively in AI systems and cloud computing. At the same time, US export restrictions on advanced AI semiconductors complicate access to the high-performance chips required for model training.
China considers artificial intelligence a strategic national priority. Beijing wants Chinese companies to develop competitive domestic AI technologies rather than relying on American firms. This creates both opportunity and pressure for companies like Alibaba.
What Investors Are Actually Worried About
The core issue is simple: Alibaba's current reporting structure makes it nearly impossible to answer basic questions. How much is the company spending on AI? Is the AI business actually profitable? How quickly is AI revenue growing? What margins do AI operations generate?
As AI becomes more central to technology companies, investors increasingly demand clearer information. Artificial intelligence businesses can appear highly profitable if infrastructure and model training costs are hidden inside broader accounting categories.
Despite concerns, Alibaba's Hong Kong-listed shares rose after the earnings report. Many investors believe the company is building a strong long-term AI position and could become one of the primary beneficiaries of China's AI expansion.
The Bigger Picture
Alibaba's next few years will likely determine whether it becomes primarily an e-commerce company with AI ambitions or one of the dominant global AI infrastructure providers. The company has advantages: massive cloud infrastructure, a large domestic customer base, strong financial resources, and government support for domestic AI development.
But success depends on whether Alibaba can monetize AI services effectively, maintain technological competitiveness, control infrastructure costs, and continue scaling cloud operations while delivering sustainable profit growth.
For IT and development professionals, this situation underscores a broader reality: the companies building the AI infrastructure you'll eventually use are investing heavily in model training, data centers, and cloud systems. Understanding those infrastructure costs-and how they're being financed-matters for evaluating both the sustainability of AI services and the competitive landscape ahead.
Learn more about Generative AI and LLM Courses or explore AI for IT & Development resources to understand how these infrastructure investments are shaping the tools and platforms you'll work with.
Your membership also unlocks:
