Targeting 100 billion USD in revenue over the next five years! Alibaba's AI organization undergoes major restructuring, positioning for the 'Token' era

Image source: Pixabay

On April 8, CEO of Alibaba Group Wu Yongming issued an internal letter announcing organizational adjustments related to AI.

It is reported that Alibaba has established a technology committee at the group level, with Wu Yongming serving as the team leader, Zhou Jingren as the Chief AI Architect of the technology committee, Li Feifei responsible for Alibaba Cloud technology and AI cloud infrastructure construction, and Wu Zemin in charge of the group’s business technology platform and AI inference platform development.

Meanwhile, Alibaba has upgraded the Tongyi Lab into the Tongyi Large Model Business Unit, led by Zhou Jingren; Li Feifei has been appointed as CTO of Alibaba Cloud; Wu Zemin will focus on his role as CTO of the Alibaba Group, while Lei Yanqun takes over as CEO of Taobao Flash Purchase.

Entering 2026, Alibaba is accelerating its AI layout.

AI organizational structure has been adjusted multiple times.

On March 2, the internal team at Alibaba Group unified the general term and core brand of AI under 'Qwen.' After the name unification, the Chinese name for Alibaba's large model brand is 'Qwen Large Model,' and the English name is 'Qwen.' 'Tongyi Lab' is the organizational name of Alibaba Group’s AI institution.

Two days later, Lin Junyang announced on a social platform his resignation as the technical lead for Alibaba’s Qwen large model. Along with Lin Junyang, several key members also announced their departures, including Hui Binyuan, head of the Qwen code direction and curator of the Qwen-Coder series models; Yu Bowen, head of post-training research for Qwen; and Kaixin Li, a core contributor to Qwen 3.5/VL/Coder, among other key personnel.

The departure of core personnel has not affected Alibaba's determination to continue its AI expansion.On March 16, Alibaba announced the establishment of the Alibaba Token Hub business group, with its core objective being 'to create Tokens, distribute Tokens, and apply Tokens,' directly overseen by Alibaba Group CEO Wu Yongming.

In an internal letter, Wu Yongming stated that the current moment is the eve of AGI (Artificial General Intelligence) explosion, where a vast amount of digital work will be supported by billions of AI Agents. These AI Agents will be powered by Tokens generated by models, becoming the main carrier for interaction between humans and the digital world.

It is reported that Alibaba Token Hub will strengthen strategic synergy across various AI businesses, reshaping workflows with AI while maintaining an agile organization. This includes: Tongyi Lab, which creates leading multimodal models and continuously pushes the upper limits of foundational model capabilities, providing the most advanced models for the group and industry; MaaS (Model as a Service) business line, which builds an efficient and open model service platform and technical system to support the entire AI ecosystem; Qwen Division, which develops the best personal AI assistant; Wukong Division, which creates a B-end AI-native work platform deeply integrating model capabilities into enterprise workflows; and the AI Innovation Division, which explores various AI innovative applications and rapidly validates new models and markets.

The day after the establishment of the Alibaba Token Hub business group, Alibaba launched the enterprise-level AI work platform 'Wukong.' Developed by the DingTalk team as a standalone application and integrated within DingTalk, it fully supports connecting users’ DingTalk accounts in enterprises, secure access permissions, and application systems. At the same time, Alibaba's ecosystem business ToB capabilities are gradually embedded in the form of skills (i.e., modular function extensions that can be installed and used without launching services). 'Wukong' will become the unified outlet for Alibaba's AI in enterprise work scenarios. Clearly, after the integration of the Qwen app into Alibaba’s C-end business scenarios, Alibaba will use Wukong to expand into B-end AI business scenarios.

This time, Alibaba established a new Group Technology Committee and upgraded the Tongyi Lab to the Tongyi Large Model Division, marking further refinement adjustments to the Alibaba Token Hub business group. In the future, more focus will be placed on leveraging advantageous resources for AI investment.

The commercial revenue target has exceeded 100 billion US dollars

In terms of performance, AI has become a growth engine.

On March 19, Alibaba announced its financial results for the quarter ending December 2025. Its total revenue was 284.843 billion yuan, representing a year-on-year increase of 2%. Adjusted EBITA reached 23.397 billion yuan, marking a 57% decrease compared to last year, mainly due to investments in instant retail, user experience, and technology. However, this decline was partially offset by improved operating performance driven by continuous growth in cloud services and increased operational efficiency across multiple businesses. Net profit attributable to ordinary shareholders amounted to 16.322 billion yuan, reflecting a 67% drop year-on-year.

Specifically, Alibaba's China e-commerce group generated revenue of 159.347 billion yuan, up 6% year-on-year; Alibaba International Digital Commerce Group reported revenue of 39.201 billion yuan, an increase of 4%; Cloud Intelligence Group earned 43.284 billion yuan, surging 36%; all other revenues totaled 67.34 billion yuan, down 25%. The Cloud Intelligence Group has now become the fastest-growing business segment within Alibaba in terms of revenue.

Regarding adjusted EBITA, the Cloud Intelligence Group also performed well. Alibaba’s China e-commerce group achieved an adjusted EBITA of 34.613 billion yuan, down 43% year-on-year; Alibaba International Digital Commerce Group recorded -2.016 billion yuan, improving by 59%; Cloud Intelligence Group reached 3.911 billion yuan, growing 25%; all other segments amounted to -9.792 billion yuan, declining by 208%.

Alibaba stated that this quarter it continued to invest heavily in two core areas: AI and consumer sectors. AI is one of the company's primary growth engines. Revenue from Alibaba Cloud grew strongly by 36%, with AI-related product revenue achieving triple-digit growth for the tenth consecutive quarter. Meanwhile, the MaaS platform business also demonstrated rapid expansion, forming a new growth driver for cloud services. On the other hand, the integration of the Qwen app with various application scenarios within Alibaba’s consumer ecosystem accelerated significantly, driving substantial new user acquisition and boosting transaction activity. The MAU of Qwen across all platforms has surpassed 300 million, entering an era of large-scale operations for AI agents.Going forward, relying on full-stack AI capabilities encompassing 'large models + cloud + chips' and comprehensive integration with Alibaba’s commercial ecosystem, efforts will continue to be made in both B2B and B2C directions for AI applications.

During the earnings call for the quarter ending December 2025, Alibaba mentioned that its AI strategy aims at a commercial goal where cloud and AI commercialization revenues, including MaaS, are expected to exceed 100 billion US dollars over the next five years.

Alibaba noted that in terms of infrastructure, Alibaba Cloud's market share has maintained growth for three consecutive quarters, increasing to 36%. By the end of February in the fiscal year 2026, Alibaba Cloud's cumulative external commercial revenue exceeded 100 billion yuan. The self-developed AI chips by Pingtouge have achieved scaled mass production. As of February 2026, they had cumulatively delivered 470,000 units on a large scale. In actual business scenarios of Alibaba Cloud, more than 60% of Pingtouge chips serve external commercial clients, completing large-scale adaptation tasks for AI purposes. The latest generation large model, Qwen3.5-Plus, excelled in comprehensive benchmark evaluations covering inference, coding, and agent intelligence. Based on innovations in underlying model architecture, Qwen3.5-Plus greatly enhanced inference efficiency while reducing costs.

For C-end applications, the monthly active users of Qwen's C-end apps have surpassed 300 million. The Qwen app has been integrated into Taobao Flash Purchase, Alipay, Fliggy, Damai, and AutoNavi, evolving into the first all-around personal assistant in China for life, work, and study.

Clearly, Alibaba's full-stack AI capabilities have become the backbone for achieving its AI strategic business objectives.

Nowadays, Alibaba is fully integrating its full-stack AI capabilities into its commercial ecosystem to accelerate the commercialization of its AI strategy. Currently, Alibaba’s Cloud Intelligence Group has seen significant revenue growth, with AI-related products maintaining high growth rates for several consecutive quarters. Alibaba has also set a target to achieve $100 billion in cloud and AI commercialization revenue within the next five years. Whether the dual focus on B2C and B2B can support Alibaba’s AI commercial goals still needs time and market validation.

Author | Wu Ren

Source | Zhengtan Finance (ID: teccj6)

7180 Views