Alibaba Eyes $100B AI Revenue Milestone Despite Plunging Q3 Profitability (BABA Q3 2026 Earnings Call)
Export as clean Markdown. Drag & drop into ChatGPT, Claude, or Gemini.
Alibaba delivered a polarizing fiscal third-quarter performance that showcased explosive growth in its cloud and artificial intelligence divisions at the direct expense of near-term profitability. Under the direction of CEO Eddie Wu, the tech giant is pivoting aggressively into the "agent-driven era" of AI, culminating in a bold pledge to surpass $100 billion in combined external cloud and AI revenue within five years. However, this unwavering commitment to future technological dominance—alongside heavy investments to capture market share in the quick commerce sector—resulted in a severe 57% contraction in total adjusted EBITA.
Cloud and AI Operations Hit Breakneck Acceleration
Alibaba's strategic pivot toward enterprise intelligence continues to bear fruit, as the Cloud Intelligence Group reported a 36% acceleration in overall revenue growth. Crucially, revenue derived from external customers surged 35%, driven primarily by an insatiable appetite for AI infrastructure. For the tenth consecutive quarter, AI-related product revenue delivered triple-digit growth, pushing the cloud division's cumulative external revenue for the fiscal year past the RMB 100 billion mark through February. The consumption of tokens on the company's proprietary Model Studio platform exploded 6x over the trailing three months, underscoring rapid adoption of its foundational large language models. To capitalize on this, the company officially centralized its various AI units under the newly established Alibaba Token Hub Business Group, serving as an organizational engine to coordinate its MaaS (Model-as-a-Service) strategy.
Generative AI Drives Consumer Integration and Chip Demand
On the product side, the company capitalized on the Chinese New Year by launching Qwen3.5-Plus and integrating its consumer-facing Qwen application across core platforms including Taobao, Alipay, and Fliggy. This integration helped drive Qwen's monthly active users past 300 million. At the infrastructure layer, the company highlighted the critical role of T-Head, its proprietary semiconductor unit. Experiencing severe global compute shortages, Alibaba revealed it has already cumulatively shipped 470,000 AI chips, capturing significant external adoption with over 60% of units supporting enterprise clients in autonomous driving and intelligent manufacturing. When questioned about a potential T-Head IPO, leadership did not rule it out, though they noted the immediate priority remains securing internal compute capacity against ongoing global supply chain constraints.
Heavy Spending Crushes Immediate Profitability
Despite a respectable consolidated revenue of RMB 284.8 billion—a 9% like-for-like increase when excluding divested assets—Alibaba's bottom line suffered under the weight of its strategic investments. Total adjusted EBITA collapsed by 57%, fueled closely by aggressive spending on both AI innovation and domestic E-commerce expansion. While the China E-commerce Group generated a solid RMB 159.3 billion in revenue, core Customer Management Revenue (CMR) crawled forward by only 1%, hampered by broader macroeconomic weakness and deep promotional discounting. However, the emerging Quick Commerce business proved to be a powerful, albeit expensive, bright spot, skyrocketing 56% to RMB 20.8 billion as Alibaba aggressively chases a RMB 1 trillion GMV target in that sector over the next two years.
E-commerce Evolves Towards Unified Agentic Workflows
Addressing concerns over shifting IT enterprise budgets, Wu outlined a fundamental thesis: corporate spending on software is migrating away from traditional siloed IT accounts and actively merging with core operational and R&D budgets as organizations deploy autonomous AI agents. Pointing to the recent launch of Wukong, Alibaba's enterprise AI agent platform, management argued that providing the infrastructure for these agentic deployments will vastly expand their total addressable market. Concurrently on the consumer side, executives noted that traditional cloud computing infrastructure must be fundamentally overhauled over the next year to support the massive memory and database demands required to host billions of constantly running, autonomous AI agents.