On January 23 Baidu reorganized two of its consumer-facing properties — its document repository (Wenku) and cloud storage service (Wangpan) — into a single business group christened the Personal Super Intelligence Business Group (PSIG). The unit is led by Wang Ying, previously head of Wenku and Wangpan, who now reports directly to chairman and CEO Robin Li, a signalling of elevated strategic priority inside China’s search giant.
The newly unified group brings together complementary assets: a massive trove of user‑generated documents and a personal cloud containing files, photos and metadata. Baidu’s internal assessment, reflected in public reporting, is that these products are relatively well positioned for AI transformation. Combined, they offer both the training data and the user touchpoints needed to deliver retrieval‑augmented models, personalised assistants and new consumer AI features.
The move is the latest step in Baidu’s broader pivot from a search‑centric business to an AI‑first company. Over the past three years Baidu has invested heavily in its large‑language model stack and generative AI services; consolidating Wenku and Wangpan creates a test bed for productising those models directly into everyday user workflows — from document summarisation and question‑answering to intelligent file search and automated content generation.
Commercially, the merger opens several monetisation paths: subscription tiers for enhanced AI features, enterprise integrations for knowledge management, and deeper conversion funnels inside Baidu’s ecosystem. It also tightens the company’s grip on the personal data flows that underpin personalised AI, giving Baidu a practical advantage in launching consumer‑facing applications that rely on private files and long‑form content.
That advantage comes with trade‑offs. Using personal and user‑uploaded content to power generative services raises privacy and compliance questions under China’s data security and personal information protection regimes. Regulators have grown more attentive to how platforms combine datasets and to the commercial reuse of personal data, so Baidu will need technical and governance safeguards alongside product innovation.
For international observers the reorganisation is a clear signal that Chinese tech firms are moving beyond model development to intensive product integration. While the domestic focus and data localisation rules limit the direct global footprint of any consumer features, the architecture Baidu is building — coupling content repositories with personal clouds and LLMs — exemplifies how companies aim to convert large, proprietary datasets into differentiated AI services.
In the near term expect faster rollout of AI‑enabled features in Wenku and Wangpan, closer integration with Baidu’s search and assistant products, and targeted monetisation experiments. Watch also for regulatory disclosures and privacy controls, which will be central to user trust and the long‑term commercial viability of Baidu’s personal‑AI play.