Baidu’s (百度) Valuation Logic Needs to Change

The angle: market lens lags Baidu’s AI turn

Chinese business outlet Huxiu argues the market is still valuing Baidu (百度) like a mature search-and-ads company, even as it refashions itself around generative AI, enterprise software, and autonomous driving. That mismatch, the commentary contends, depresses the company’s multiples and obscures new revenue pathways. Is Baidu still an advertising play, or has it become an AI infrastructure and applications platform?

From search cash cow to AI platform bets

Baidu has leaned hard into foundation models and applied AI, integrating its ERNIE large language model into search, smart devices, and cloud offerings. In China’s enterprise market—shaped by state digitalization drives and procurement norms—such AI “stacks” can ride long sales cycles but offer recurring revenue potential. Autonomous driving and robotaxi services under Baidu’s Apollo brand remain a high-profile long game; commercialization is gradual, but reportedly expanding pilots and partnerships suggest continued policy support in select cities.

Valuation headwinds: geopolitics and regulation

The rerating case runs into familiar China tech headwinds. Since 2020, platform regulation, a property-led growth slowdown, and U.S.–China tensions have pressured investor sentiment and discounted earnings visibility across the sector. Washington’s export controls on advanced AI chips restrict access to leading GPUs, complicating model training and inference economics for Chinese AI developers. Baidu has invested in domestic accelerators and optimization, but capacity, efficiency, and cost remain watch points. Overlay this with ADR geopolitics, audit scrutiny, and evolving rules for generative AI in China—where interim measures set content and security baselines—and investors still demand a margin of safety.

What could force a rerating?

Huxiu’s thesis is that Baidu should be valued less like a cyclical ad vehicle and more via sum-of-the-parts: a cash-generating search core, an AI cloud and enterprise software arm, and long-dated autonomy options. Catalysts? Clear monetization of generative AI in search and enterprise workloads, evidence of sustainable AI cloud growth, and credible unit-economics milestones in robotaxis. Absent that, multiples may stay tethered to legacy perceptions. The broader question for Western investors is simple: when do China’s policy, chip-supply, and demand variables stabilize enough for the market to price Baidu as the AI company it wants to be?