Baidu launches a new round of offensive

Author | Zhou Zhiyu

Editor | Zhang Xiaoling

A transformation in AI applications has reached a critical point.

Wall Street Insights has learned from informed sources that Baidu has recently officially announced the establishment of the Personal Super Intelligent Group (PSIG). Sources close to Baidu revealed that PSIG will integrate core AI application assets such as cloud storage and document library, led by Vice President Wang Ying, and will report directly to CEO Li Yanhong.

Cloud storage and document library originally belonged to the Mobile Ecosystem Group (MEG), which means that in addition to MEG, the Intelligent Cloud Group (ACG), and the Intelligent Driving Group (IDG), Baidu has added another business group. Sources close to Baidu indicated that the overall trend for cloud storage and document library is currently good and has received attention from Li Yanhong. From a management perspective, the document library and cloud storage have long been merged under Wang Ying's oversight.

In the past, AI applications in Baidu's landscape were often seen as a "showroom" to showcase technological prowess to the capital market, and there were doubts about when these technologies could be transformed into substantial revenue. Baidu's bold adjustments this time clearly set higher expectations for the market of AI applications.

In a sense, PSIG will be an independent strategic support point for Baidu's next decade. It will provide a positive answer from Baidu on how AI can transform into productivity and be reflected in financial reports.

By merging cloud storage (personal asset management) with the document library (intelligent content creation), Baidu aims to compensate for the growth shortcomings of a single search tool and leverage its strengths in processing existing user data.

Of course, in this trillion-level AI race, Baidu never lacks competitors. ByteDance's Doubao has formed a hunting momentum at the C-end entry; Alibaba's Tongyi series has made deep investments in office scenarios; Tencent's Yuanbao, backed by its social ecosystem, is determined to seize the high ground of large model applications by 2026.

Industry insiders told Wall Street Insights that the positioning of AI applications is to complement traditional search. The advantages of search have already been established, but "if the traffic logic completely changes, you must have a capable application entry; you cannot let competitors take control."

Moreover, the demand for productivity growth driven by AI has brought new opportunities for Baidu. Computing power is not only a major cost but also an "efficiency engine." In an environment where traditional search advertising is under pressure, the stability of pure traffic distribution has deteriorated, and there are even growth bottlenecks. However, the AI processing capabilities based on the document library and cloud storage have far greater commercial value than traditional keyword clicks.

At the same time, users' demands for AI assistants are "immediate feedback." The high stickiness of Baidu's PSIG allows it to "connect in that moment" when user needs arise, ensuring that user assets do not flow to competing platforms.

Although Baidu's PSIG capacity has not been fully released, products under it, such as the document library, have already proven to have strong commercialization potential. Baidu's document library currently has over 1.8 billion authoritative professional documents, with over 97 million monthly active AI users; Baidu Cloud Storage serves over 1 billion users, with over 80 million monthly active AI users.

This is a smoke-free hunting ground. Everyone is betting that 2026 will be the "industrialization year" for AI-native applications. Whoever can first connect user experience and business models will hold the ticket for the next decade UBS Securities China Internet industry analyst Xiong Wei believes that by 2026, we should see a gradual enrichment of overall AI application scenarios on the application side. The scenarios will become more diverse, and from a commercialization perspective, there are opportunities for further acceleration.

Currently, the next upward cycle in the internet industry is primarily driven by AI applications. For applications to explode, they must provide an exceptional experience. When traffic costs rebound, only "super intelligence" can act as a "growth stabilizer." This also means that as the urgency of the current revenue structure transformation rises, Baidu hopes to have more users and investors accept its determination for AI transformation at this time.

Over the past decade, search advertising has been the absolute king. Its fluctuations dominate the rise and fall of companies with market capitalizations in the hundreds of billions. However, Baidu's establishment of PSIG, led directly by Wang Ying, is a proactive adjustment in response to the changing times.

As users become more accepting of AI assistants, the "thrilling leap" from testing demos to large-scale products is just a matter of time.

This means that internet competition after 2026 will no longer be a single-dimensional battle for traffic but will evolve into an ecological balance where algorithms, data, and scenarios coexist diversely. This will create a more resilient business system. In this system, growth will no longer be easily out of control, and valuations will no longer be fragile.

AI-native applications resolve the scarcity of traffic dividends, while the certainty of organizational structure adjustments hedges against the uncertainties of future changes. Major internet companies are each taking action to respond to the AI upheaval