Key Takeaways
- Baidu has restructured MEG, merging its Commerce Department and E-Commerce Business Unit into a new Big Commerce Business Unit and elevating its digital human division to an independent department
- The reorganization is part of a broader shift to rebuild search, advertising, and commerce around AI agents
- For e-commerce operators, it offers a window into how search platforms themselves are turning into the entry point for commerce
Baidu's Commerce Consolidation Toward an AI-Native Model
Baidu's Mobile Ecosystem Group (MEG) has undergone another significant organizational shakeup, consolidating its commerce and e-commerce operations into a single business unit as the Chinese internet giant accelerates its transition toward an AI-native model.
pandaily.comIn June 2026, Chinese search giant Baidu reshuffled its Mobile Ecosystem Group (MEG) once again. The restructuring merges the former Commerce Department and the E-Commerce Business Unit into a newly formed Big Commerce Business Unit, led by Vice President Ping Xiaoli.
What stands out is that, alongside this merger, the Digital Human Innovation Business Unit was upgraded to an independent department, also placed under Ping's leadership. The intent is to bring advertising, content, commerce, and digital human monetization together within a single AI-first framework.
Ping joined Baidu as an intern in 2007 and later led the Baidu App, growing its daily active users from 100 million to 230 million. She was appointed a vice president in August 2025, and this assignment underscores how much weight the company now places on commerce.
The Context of 2026's Rapid Reorganizations
This MEG restructuring is not an isolated event. Since the start of 2026, Baidu has carried out one organizational change after another to bend its entire product portfolio toward AI.
Tracing the timeline makes the intent clearer. In January, Baidu merged Wenku (Docs) and Wangpan (Cloud Drive) into a new Personal Super Intelligent Group (PSIG), reporting directly to CEO Robin Li. In March, it integrated its large language models into search and recommendation products. In May, it set up the Baidu Model Committee (BMC) to unify LLM R&D, bringing in younger AI researchers to shorten decision cycles. The June MEG restructuring is the latest phase of this transformation.
Behind this run of reorganizations lies an urgency to rebuild search — Baidu's core franchise — with AI. According to kr-ASIA, Baidu is embedding ERNIE into the Baidu App to deliver LLM-powered search to 700 million users. If the entry point to search moves to AI, the commerce layer beyond it naturally has to be rebuilt as well — a continuous line of logic.
What the Digital Human Bet Actually Contains
The most symbolic part of this restructuring is the independence granted to the digital human business. Baidu is positioning this area at the heart of its commerce strategy.
At the center is the Yijing platform, formerly known as Huiboxing. The company describes it as the world's first full-scenario, multi-agent digital human platform. A digital human refers to a virtual persona generated and operated by AI — here, used as a livestream host (anchor) for live commerce. By the end of 2025, Yijing had onboarded over 100,000 digital human anchors operating across more than 30 industries.
That this bet is no abstraction is borne out by results. In June 2025, an AI avatar of well-known influencer Luo Yonghao made its debut livestream on Baidu's e-commerce platform. According to TechNode, it drew over 13 million viewers and generated roughly 55 million yuan (about $7.66 million) in gross merchandise value (GMV).
The technology behind those numbers is just as striking. Over the six-hour session, the AI system drew on a 13,000-item product knowledge base, generated 97,000 words of product descriptions, and coordinated more than 8,300 avatar movements. Deploying large numbers of digital humans that can run around the clock offers more economies of scale than hiring a single top human host — and Baidu clearly senses traction in this direction.
From Search to Agents: Baidu's Strategic Coordinates
Why is Baidu placing such emphasis on commerce and digital humans? The answer lies in how the company is redefining itself from a "search company" into an "agent company."
In early 2026, CEO Robin Li used an internal address to clearly categorize Baidu's agents into four types. The top priority is the "search agent," followed by "digital human agents" under the Huiboxing brand, "coding agents" distributed through Miaoda, and others. At the Create 2026 event, Li further proposed "Daily Active Agents" as a new metric for evaluating AI product output.
What matters here is that digital humans are framed not as mere broadcasting tools but as a category of agent. They understand product knowledge, respond to viewer questions, and steer toward purchase — a form of agentic commerce that completes transactions through conversation. A division of labor emerges: search agents discover products, while digital human agents handle persuasion and purchase.
The Bigger Picture of Chinese Platform Competition
Baidu's moves need to be understood as part of the agentic commerce race that China's tech giants are advancing in unison. Each is replacing the search bar with AI agents and casting commerce as the new battleground.
The intensity of the competition shows up in spending. According to reports including CNBC, Chinese giants spent roughly $647 million subsidizing AI shopping adoption during the 2026 Lunar New Year alone — Alibaba ($431M), Tencent ($144M), and Baidu ($72M).
This clarifies Baidu's position. Alibaba has reached 300 million monthly users across its consumer platforms with Qwen at the core, while Tencent is embedding agent capabilities into WeChat, with its 1.3 billion users. Against these two — both armed with super apps and ecosystems — Baidu differentiates with its own strengths in search and digital humans. The MEG restructuring reads as a move to concentrate limited resources on these two weapons.
| Platform | Core agent foundation | Central commerce approach |
|---|---|---|
| Baidu | Search agent / ERNIE | Search-led discovery plus digital human broadcasting |
| Alibaba | Qwen | AI shopping integrated into consumer platforms |
| Tencent | In-WeChat agents | Purchase, booking, and payment inside the super app |
What E-Commerce Operators Should Take Away
For e-commerce operators outside China, a distant corporate reshuffle might seem only loosely relevant. Yet the structural shift underway here carries meaning that transcends any single market.
The biggest takeaway is that search platforms themselves are turning into the entry point for commerce. Users ask an AI, the AI discovers, compares, and recommends products, and a digital human steers them toward purchase. Within this flow, the traditional model of placing ads on a search results page is becoming transitional. Designing so that product information is correctly understood and recommended by AI agents is becoming a new axis of competition.
The other takeaway is the possible shift of live commerce hosts from humans to AI. As the Luo Yonghao avatar case shows, in a world where digital humans trained on top hosts' abilities run 24 hours a day, the cost structure and scale of broadcasting change fundamentally. Building out a product knowledge base — preparing the data so that AI can describe your products accurately — becomes a more important area of investment than ever.
Conclusion
Baidu's MEG restructuring is one cross-section of a search company molting into a commerce company, and further into an agent company. The merger of commerce and e-commerce, and the independence of the digital human business, both express a clear intent to redesign monetization around AI.
Leading the way in the Chinese market, this move foreshadows a future where the line between search and commerce dissolves. In a world where AI agents handle everything from discovery to purchase, how your products get described will matter. Preparing to face that question is exactly what is now being asked of every operator.
