In a decisive move to streamline its artificial intelligence initiatives and accelerate commercialization, Alibaba Group has established a new business unit dubbed “Token Foundry.” The restructuring, which consolidates key AI research teams under the direct supervision of CEO Eddie Wu Yongming, underscores the e-commerce giant’s aggressive push to dominate China’s fiercely competitive generative AI landscape.
The formation of Token Foundry, officially launched on June 8, 2026, merges Alibaba’s Tongyi Lab with the Future Life Lab. This consolidation brings together the company’s foundational large language model (LLM) development and its cutting-edge consumer AI applications into a single, cohesive entity.
Centralizing AI Leadership
The restructuring represents a significant centralization of power within Alibaba’s Big Tech hierarchy. By placing Token Foundry under his direct leadership, CEO Eddie Wu is signaling that artificial intelligence is no longer just a supporting technology for Alibaba’s e-commerce and cloud divisions, but the core driver of the company’s future growth.
According to a report by the South China Morning Post, the Future Life Lab was previously housed within the Taobao and Tmall Group. Under the guidance of Zheng Bo—chief scientist and vice president of technology at Taobao and Tmall Group, as well as CTO of Alimama—the lab developed innovative consumer-facing AI products, including the video-generation model “Happy Horse” and the open-world AI model “Happy Oyster.”
The integration of these consumer applications with the foundational research of Tongyi Lab is designed to shorten the development cycle from core model training to user-facing product deployment. This agility is critical as Alibaba competes with rivals like Tencent, Baidu, and a slew of well-funded startups in the race to capture the domestic AI market.
Shuffling the Scientific Deck
The creation of Token Foundry is accompanied by significant leadership changes among Alibaba’s top scientific talent. Zhou Jingren, who played a pivotal role in building Tongyi Lab and overseeing the development of the widely adopted Qwen family of LLMs, has been elevated to the position of Alibaba chief scientist.
In his new role, Zhou will head the newly formed Alibaba AI Future Research Institute, which will focus on frontier AI research and next-generation breakthroughs. Zhou’s promotion to the Alibaba Partnership, the company’s highest decision-making body—in late 2024 highlighted his growing influence within the organization.
Meanwhile, Zhang Di, who previously led the development of Kuaishou’s Kling video model before rejoining Alibaba in November 2025, will continue to run the Future Life Lab under the new Token Foundry umbrella. This concentration of top-tier talent reflects Alibaba’s commitment to maintaining its edge in multimodal AI capabilities, particularly in the rapidly evolving video generation space.
The Commercial Imperative
Token Foundry will operate under the broader Alibaba Token Hub (ATH), which was established in March 2026 to oversee the company’s model-as-a-service offerings, the Qwen consumer app, and enterprise AI products. This structure aligns with Alibaba’s dual strategy of providing foundational AI infrastructure to enterprise clients through Alibaba Cloud while simultaneously developing proprietary AI applications for consumers.
The commercial imperative behind this restructuring is clear. In its first-quarter 2026 earnings report, Alibaba posted revenue of 243.4 billion yuan ($33.9 billion), slightly missing the Bloomberg consensus estimate of 246.5 billion yuan. However, the bright spot was Alibaba Cloud, which saw revenue surge 38 percent year-on-year to 41.6 billion yuan, driven by a 40 percent increase in external cloud customers.
During the earnings call, CEO Eddie Wu emphasized the explosive growth of the company’s AI services. He projected that annualized recurring revenue from AI models and application services would exceed 10 billion yuan in the second quarter and reach an ambitious 30 billion yuan by the end of the year.
Qwen’s Global Footprint
Central to Token Foundry’s mandate is the continued development and expansion of the Qwen family of large language models. Qwen has become one of the most widely adopted open-source AI model series in the world, accumulating over 600 million downloads on Hugging Face and spawning more than 170,000 derivative models created by developers across the globe. This extraordinary adoption rate has made Qwen a de facto standard in the open-source AI community, rivaling Meta’s Llama series in terms of global reach.
The Qwen models are also central to Alibaba’s international cloud strategy. By offering Qwen through Alibaba Cloud’s API services, the company is competing directly with OpenAI and Anthropic for enterprise clients in Southeast Asia, the Middle East, and Europe. The model’s strong multilingual capabilities, particularly in Chinese, Arabic, and other non-Western languages, give Alibaba a distinct competitive advantage in markets that Western AI companies have historically underserved.
A Unified Strategic Vision
To ensure alignment across its sprawling empire, Alibaba formed a group-level technology committee in April 2026. Chaired by Wu, the committee includes chief scientist Zhou Jingren, group CTO Wu Zeming, and Alibaba Cloud CTO Li Feifei. This high-level coordination is essential as Alibaba seeks to integrate AI capabilities across its e-commerce platforms, logistics networks, and digital entertainment divisions.
The creation of Token Foundry also comes at a moment when Alibaba is navigating significant geopolitical headwinds. The company’s recent addition to the US Department of Defense’s Section 1260H list of Chinese military companies has added reputational risk for international clients. However, Alibaba’s leadership appears to be doubling down on its AI strategy rather than retreating, betting that the commercial value of its models and cloud services will ultimately outweigh the political complications.
The launch of Token Foundry is a clear indication that Alibaba is moving past the experimental phase of AI development and into aggressive commercialization. By breaking down internal silos and consolidating its top AI talent under the direct control of the CEO, Alibaba is positioning itself not only to compete in the current AI price wars but also to define the next generation of intelligent applications in China and beyond.