Baidu’s AI Business Surpasses Search Advertising for the First Time: The Pros and Cons of Structural Transformation

The Decline of the Search Advertising Era

Baidu’s Q1 2026 financial results have marked a historic turning point in its business structure. Revenue from online marketing services fell 22% year-on-year to 12.6 billion yuan, accounting for only 48% of total revenue—a drop below the majority for the first time.

Meanwhile, revenue from Baidu’s new AI businesses, centered around intelligent cloud infrastructure, reached 13.6 billion yuan, surpassing half of the total revenue. Robin Li, Baidu’s founder, stated, “AI has become the core driving force of Baidu,” emphasizing the role of growing corporate demand and the company’s long-standing full-stack AI capabilities in driving growth.

Tug-of-War Between Old and New Businesses: The Challenges of Structural Transformation

While this milestone highlights Baidu’s evolution, it also reveals some tough realities. Revenue from Baidu’s “traditional businesses” dropped by 29% year-on-year to 10.2 billion yuan. The decline stems from two major pressures: first, competition from content platforms like TikTok, Kuaishou, and RED, which continue to attract advertising budgets; and second, the rise of AI search technology that shifts user behavior from clicking paid links to directly obtaining answers. Baidu’s own Ernie AI search innovations are accelerating this transition.

There are, however, some positive signs. Revenue from AI-native marketing services grew 36% year-on-year to 2.3 billion yuan, suggesting that advertisers are beginning to adjust their budgets to align with changes in traffic and product formats. The painful transition from traditional bidding-based rankings to AI-native integrated solutions mirrors the logic of the shift from portal advertising to search advertising two decades ago. Just as Baidu emerged victorious during that prior transition, the company now faces another opportunity to prove its resilience.

Driving Growth: The Rise of Intelligent Cloud

The standout highlight from this quarter’s financial results was the performance of Baidu’s intelligent cloud infrastructure business. Revenue soared to 8.8 billion yuan, a remarkable year-on-year growth of 79% and a quarter-on-quarter increase of 52%. Particularly noteworthy was the 184% year-on-year growth in GPU cloud revenue, setting a benchmark within the financial report.

On the hardware side, Baidu’s proprietary Kunlun chip P800 successfully completed large-scale testing, delivering multiple clusters with over 10,000 cards. These chips reportedly meet the requirements for training advanced large-scale AI models. The next-generation supernode based on Kunlun chips, the Tianchi 256-card system, has also been launched and is set for release in June. Baidu is leveraging its full-stack strengths—including foundational AI computing power and end-to-end solutions based on the Ernie large-scale model—to strengthen its appeal to enterprise customers.

Challenges in AI Applications and Autonomous Driving

Baidu’s AI applications generated 2.5 billion yuan in revenue, remaining flat year-on-year but declining by 10% compared to the previous quarter. Despite recent launches of new products like the DuMate agent and Miaoda 3.0 no-code platform, the commercialization of AI products remains a slow process, reflecting a shared challenge across the industry.

Meanwhile, Baidu’s autonomous taxi service, Apollo Go, showed promising growth, with over 3.2 million fully driverless ride orders—a year-on-year increase of more than 120%. The service has expanded operations to Dubai, with plans to enter the London market. Its cumulative driving distance has exceeded 330 million kilometers, making it the world’s largest commercial robotaxi data repository.

However, challenges have arisen. On March 31, over 100 Apollo Go vehicles in Wuhan simultaneously stopped operating due to a system failure. While Baidu officially attributed the incident to technical issues, industry analysts suggest that the heavy reliance on vehicle-cloud collaborative architecture may have led to a large-scale operational halt. This event has exposed vulnerabilities in emergency response mechanisms for large-scale autonomous driving operations, highlighting significant commercialization challenges.

Cost Management and Future Growth Synergies

Baidu showed improvement in profitability through proactive cost control measures. Non-GAAP operating profit rose 39% quarter-on-quarter to 4 billion yuan, as research and development expenses and selling, general, and administrative costs fell by 22% and 34%, respectively. This indicates a strategic concentration of resources in core areas such as cloud infrastructure and large-scale AI models. With 280 billion yuan in cash and investments, Baidu’s management has emphasized its ability to make calculated investments for future growth.

However, its video streaming service subsidiary, iQIYI, remains a financial drag, reporting revenue of 6.2 billion yuan against an operating loss of 228 million yuan. The lack of clear synergy between iQIYI and Baidu’s core AI business has become a point of concern for market analysts.

Baidu’s latest financial results underscore the steady progress of its AI-led structural transformation but also highlight several challenges. These include the decline of traditional businesses, the extended timeline for commercializing AI applications and autonomous driving, and the need to integrate various investment areas to generate synergistic growth. The key to Baidu’s future lies in how effectively it can leverage explosive growth in intelligent cloud services, along with its Ernie large-scale model, AI-native marketing, Apollo Go, and agent ecosystems, to drive sustainable overall growth.