Alibaba Bets Big on AI Agents with Token Hub and Wukong Platform

Key Highlights of the News

• Alibaba has reorganised its artificial intelligence operations by creating a new division called the Alibaba Token Hub (ATH).
• The new unit will be led directly by CEO Eddie Wu and aims to unify the company’s AI efforts.
• ATH brings together major AI assets including Tongyi Laboratory, the Qwen model and multiple innovation teams.
• The restructuring separates AI operations from Alibaba’s cloud business, signalling a shift in strategy.
• Alibaba has launched a new enterprise AI platform called Wukong to automate complex business tasks.
• Wukong allows multiple AI agents to work together for tasks such as document editing, data processing and research.
• The platform is currently in invitation-only beta and integrates with DingTalk and other enterprise tools.
• The move comes amid rising competition in China’s AI market, especially around AI agent-based systems.

Alibaba Group has undertaken a significant restructuring of its artificial intelligence operations while simultaneously introducing a new enterprise-focused AI platform, as it steps up efforts to strengthen its position in the rapidly evolving AI sector.

As part of this restructuring, the company has established a new business division named the Alibaba Token Hub (ATH), which will be led directly by Chief Executive Officer Eddie Wu. The objective of this move is to bring together Alibaba’s various AI initiatives under a unified organisational framework, enabling faster development and better coordination across its AI ecosystem.

The newly formed unit consolidates several of Alibaba’s core AI assets, including its research arm Tongyi Laboratory, the Qwen large language model and multiple AI-related business and innovation teams. By integrating these components into a single structure, the company aims to streamline its AI strategy and improve execution efficiency.

This restructuring also reflects a broader strategic shift. Alibaba is separating its artificial intelligence operations from its cloud computing business, indicating a move away from its earlier integrated “full-stack” approach. The change suggests a stronger focus on developing independent AI capabilities and building a more clearly defined commercial model around them.

Industry observers view this reorganisation as an effort to create a sustainable revenue model for artificial intelligence. The company is expected to increasingly adopt a token-based system, where AI services are monetised based on usage. This approach is particularly relevant for AI agents, which operate continuously and execute multi-step tasks, resulting in significantly higher computing consumption compared to traditional chatbot-based systems.

Alongside the organisational changes, Alibaba has launched Wukong, a new AI platform designed specifically for enterprise use. The platform focuses on automating business workflows by enabling multiple AI agents to operate together within a single interface. These agents can handle a wide range of tasks, including document editing, spreadsheet updates, meeting transcription and research activities.

Wukong is currently available in an invitation-only beta phase. It can be accessed either as a standalone application or through DingTalk, Alibaba’s workplace collaboration platform used by millions of corporate users. The company also plans to integrate the platform with other widely used enterprise communication tools, further embedding AI capabilities into everyday business operations.

These developments come at a time of intensifying competition in China’s artificial intelligence market. Technology companies across the country are rapidly introducing AI agent-based solutions, driven in part by growing interest in systems capable of automating complex workflows. At the same time, the rapid adoption of such technologies has raised concerns related to security, cost structures and long-term profitability.

Overall, Alibaba’s latest initiatives highlight its transition toward an AI-first strategy, with a clear focus on enterprise automation and scalable AI services. The combination of organisational restructuring and new product launches reflects the company’s effort to position itself more effectively in a competitive and fast-changing AI landscape.