Key Points:
- Alibaba is shifting its AI focus to “agents” that connect its various businesses.
- The new Alibaba Token Hub business group signals this move towards more complex AI.
- Alibaba’s existing wide range of services gives it an edge over competitors in integrating AI agents.
- The company aims to monetize AI agents, which consume more data than simple chatbots.
Alibaba is sharpening its AI strategy, focusing on “agents” that link its many businesses together. These AI agents are like smart assistants that can do more than just answer questions; they can help users make purchases and manage complex tasks across Alibaba’s vast empire.
In recent months, Alibaba has introduced several AI agent features. This week, the company announced it would separate its AI businesses from its cloud division, creating the new Alibaba Token Hub business group. This move, led by CEO Eddie Wu, shows Alibaba’s clear intention to focus on digital assistants powered by advanced AI models that use much more data than traditional chatbots.
Alibaba, a $325 billion e-commerce giant, will release its quarterly results on Thursday. Everyone is watching to see how the company plans to make money from AI, a challenge many tech companies face. Analysts expect Alibaba’s third-quarter revenue to go up by 3.8%, but its net income to drop by 42.5%. This quarter included Singles’ Day, China’s biggest shopping event.
With Chinese shoppers saving more and spending less due to economic worries and a housing crisis, Alibaba is looking for new ways to encourage buying. Last year, it invested heavily in its instant retail platform, competing with Meituan for one-hour deliveries. This year, Alibaba’s AI chatbot, Qwen, has started to help users make direct purchases through chat.
An early attempt to get users to try Qwen’s new shopping features in February faced a snag. Alibaba launched a 3 billion yuan ($435.7 million) coupon campaign allowing in-app purchases using only chatbot prompts. The coupons were so popular that the app temporarily shut down.
Brian Wong, a former Alibaba employee and author, believes that integrating daily tasks across Alibaba’s huge ecosystem – which includes e-commerce, food delivery, travel, and more – into a chatbot could fundamentally change how people shop. He compares it to having many major platforms all working together through a simple text box. He thinks this kind of integration is happening first in China because of the challenges of combining different company platforms in places like the U.S.
Alibaba isn’t the only Chinese tech company using AI agents for consumer tasks, but its rivals like Tencent and ByteDance would mostly act as platforms for third-party companies. Alibaba’s own ecosystem gives it a big advantage, according to analyst Ed Sander. He points out that Alibaba handles everything from the chatbot to logistics and runs it all on its own cloud, a capability no other company has.
On Tuesday, Alibaba also launched Wukong, an enterprise-focused AI platform for automation. Wukong can manage multiple AI agents to handle complex business tasks like document editing and meeting transcriptions within one interface.
A major reason for this shift to AI agents is not just the excitement around new AI technologies but also the potential for profit. These agents can make decisions and work around the clock, consuming much more data than a typical chat session. This is important for Chinese companies, many of which offer free open-source AI models, leading to falling data prices due to intense competition.
This AI push comes as Alibaba faces leadership changes in its AI division. Lin Junyang, head of the Qwen model division, left in early March, becoming the third senior Qwen executive to leave this year. Morningstar analyst Chelsey Tam noted that this raises concerns about morale and Alibaba’s ability to keep top AI talent, which is scarce. However, the company’s cloud division has enough talent to fill any gaps.
