Baidu World 2026: Robin Li Proposes "Intelligent Organization" and Three Principles of Organizational Evolution
At Baidu World 2026, Robin Li proposed that the next act of AI is the intelligent organization, and for the first time put forward the "Three Principles of Organizational Evolution," advocating that agents and digital employees enter core business areas such as sales, customer service, R&D, and operations. Within Baidu, the number of daily active agents has grown from 5,000 to 37,000, the number of effective tasks completed per week has increased from 84,000 to over 1 million, and 4,300 Workspaces have been built, accumulating over 20,000 organizational Skills. Baidu's Dazi has been upgraded with the Harness engine, Kuku AI has shifted toward multi-expert collaboration, the overseas version Kooko has surpassed 10 million users, and the "Token Institute" for enterprises and developers has been released.
Body · AI translation
Baidu World 2026: Robin Li Bets AI's "Next Act" on the Intelligent Organization
Thirty-seven twenty-one, 33 minutes ago. As AI enters companies, how must organizations change?
Ten years ago, at Baidu World 2016, Robin Li proposed that "the next act of the internet is artificial intelligence." Ten years later, he believes that as the "chip-cloud-model-agent" technology architecture—comprising chips, cloud, large models, and agents—gradually matures, what AI will change next is the organization itself.
Image source: Baidu
On the morning of November 12, Baidu World 2026 was held in Beijing. In his keynote speech, Baidu founder Robin Li proposed that the next act of artificial intelligence will be the intelligent organization, and for the first time put forward the "three principles of organizational evolution."
"If artificial intelligence is the new electric power, then the intelligent organization is the new factory of the AI era—the place where productivity truly happens," Robin Li said. Companies must not treat AI merely as a personal efficiency tool; they must also bring agents and digital employees into core businesses such as sales, customer service, R&D, and operations, and redesign organizations and workflows around AI.
Image source: Leitech
Why do companies invest more and more in AI, yet revenue and profit do not necessarily grow in tandem? In his speech, Robin Li raised this problem that many companies are facing. In his view, the reason is that many companies have only replaced the "motor" without redesigning the "factory." To this end, he proposed three principles of organizational evolution:
Robin Li also predicted that jobs will not disappear in the future, but will gradually converge into two major types: Builders, responsible for creating new things, and Sellers, responsible for communication, connection, and building trust.
At the event, Robin Li also spoke with Lenovo Group Chairman and CEO Yang Yuanqing about the changing role of corporate managers in the AI era. Yang Yuanqing noted, "AI doesn't understand customers at all; this emotional connection between people is still very important, so on this business trip I also met quite a few customers, and analysts held quite a few roundtables."
Image source: Leitech
Meanwhile, Baidu is also conducting corresponding experiments internally. According to disclosures by Robin Li, over the past year, the number of daily active agents within Baidu grew from 5,000 to 37,000, and the number of effective tasks completed by agents each week grew from 84,000 to more than 1 million.
At present, Baidu has established 4,300 Workspaces internally and accumulated more than 20,000 organizational Skills, enabling employees' experience and working methods to be reused by others and invoked by agents.
On the product side, the focus of this upgrade to Baidu Dazi is professional tasks. The new Harness engine is oriented toward long-chain work; a three-layer memory and self-evolution system can retain task experience; and the enterprise collaboration space allows employees and agents to share project context, decisions, and outputs. Baidu has also partnered with others to launch professional workbenches for teaching, video creation, and sales, further connecting industry software with workflows.
Kuku AI's upgrade centers on "from individual to organization," emphasizing capabilities such as multi-expert collaboration, device-cloud relay, and Computer Use. It can invoke the public professional content of Wenku and also combine users' own files to complete deliverables such as research reports, data analysis, and PPTs. The enterprise edition further connects to Netdisk Enterprise Edition and team spaces, making organizational files, data, and collaboration relationships the working context for agents.
In addition, overseas Kooko users had previously surpassed 10 million, and Baidu has thereby unified Kuku AI's global brand.
At the conference, Baidu also officially launched "Ciyuansuo" for enterprises and developers, allowing users to compare information such as prices, response speeds, and success rates across different models and different service providers before making purchases and API calls. Addressing the current problem of uneven quality in model relay services, Baidu has also introduced model "verification" and supplier "credit checks," hoping to turn model services into a more transparent trading market.
More Skills, agents, and product features do not mean organizational efficiency has improved by much.
From last year's proposal to "internalize AI capabilities" to this year's further emphasis on organizational evolution, Baidu's discussion of AI's value is shifting from personal efficiency to how enterprises work. However, the real challenge is still how to enable reliable collaboration among agents across different roles and different systems, and how much business benefit these changes can actually bring.
For the "intelligent organization," these questions are probably more important than how many more agents are added.
Images in this article are from: 123RF licensed image library. Infringement report. Back. Share. Related articles. All comments.