Alibaba Creates Token Hub Group to Unite AI Development Under CEO Eddie Wu
Alibaba Group Holding has reorganized its artificial intelligence operations by launching a new business division called the Alibaba Token Hub Business Group. The move places all major AI teams and products under one centralized structure led directly by CEO Eddie Wu Yongming. The Hangzhou based technology company said the initiative is designed to strengthen coordination between its growing portfolio of AI projects as the global competition around large language models intensifies. The Token Hub will operate alongside Alibaba Cloud and the company’s core e commerce businesses, reflecting how central artificial intelligence has become to the firm’s future growth strategy.
The newly created group gathers several important research and product teams that previously operated in different areas of the company. These include Tongyi Laboratory, which develops the Qwen series of large language models, as well as the company’s model as a service platform that allows developers and businesses to deploy AI models. The restructuring also brings together the Qwen unit focused on personal AI assistants and a new enterprise technology group named Wukong. The Wukong team will work on integrating artificial intelligence into workplace collaboration tools built around the Dingtalk business platform used by many companies across China.
Executives say the Token Hub is built around the concept of tokens, which represent the basic units generated by AI models when they process or produce content such as text, images or programming code. These tokens form the foundation of how large language models calculate and generate results. By organizing development around this principle, Alibaba hopes to accelerate both model innovation and real world applications. According to Wu, the objective of the Token Hub is to create tokens, deliver tokens and expand the use of tokens across business and consumer services. The company believes this framework can unify research, engineering and product deployment within its AI ecosystem.
Wu also told employees that the artificial intelligence industry is approaching an important turning point. Rapid improvements in large scale models are expected to make it possible for billions of intelligent software agents to handle increasing portions of digital work. These agents could automate tasks ranging from writing documents and analyzing data to coordinating complex enterprise processes. Within Alibaba’s vision, tokens will serve as the underlying economic and computational units enabling this expanding network of AI services. The company sees strong potential for AI driven productivity tools integrated with its cloud infrastructure and enterprise platforms.
The restructuring follows a period of internal changes within Alibaba’s AI teams. Earlier in the month several researchers associated with the Qwen model project left Tongyi Laboratory, including technical lead Lin Junyang and other senior engineers. Talent movement is common in the fast evolving AI sector where experienced engineers are in high demand across technology companies and startups. To maintain momentum in foundational model research, Wu informed staff that a dedicated task force would coordinate resources from across the company to speed up development of advanced AI systems and strengthen collaboration between research and product teams.
Alibaba has also continued recruiting global AI talent as it expands its research capabilities. The company recently hired Zhou Hao, a former senior research scientist at Google DeepMind, to lead work on post training methods for large models. Post training refers to the process of refining AI models after their initial training stage to improve reasoning, reliability and performance. Strengthening this area has become a priority for major AI developers as they seek to enhance the practical capabilities of large language models used in business applications, consumer assistants and enterprise software platforms.
The launch of the Token Hub group marks another step in Alibaba’s ongoing transformation as it adapts to a rapidly changing technology landscape shaped by artificial intelligence. By centralizing its most important AI teams under one leadership structure, the company aims to streamline innovation and compete more aggressively with global technology leaders building advanced model ecosystems. Artificial intelligence is increasingly becoming a core pillar of Alibaba’s strategy as it seeks to expand its influence across cloud computing, enterprise services and next generation digital platforms powered by intelligent software systems.
