Baidu’s Q2 AI Pivot Is Growing—but Advertising and Profit Are Still Falling

Baidu’s second-quarter 2026 performance captures the central tension in its transformation: its AI businesses are expanding quickly, but the company’s traditional advertising engine is contracting faster than the new businesses can replace it. Total revenue fell 4% year over year to RMB31.33 billion, below the RMB31.96 billion analyst consensus, while net profit fell 68% to RMB2.32 billion. Revenue has now declined for more than a year, and profit has more than halved for a third consecutive quarter. 21011

The legacy business is still setting the pace

Online marketing services, Baidu’s traditional financial foundation, generated RMB13.1 billion in the quarter—down 19% from a year earlier. The company cited intensifying competition for user attention and a deliberate delay in monetizing AI-powered search, while weaker advertiser demand added to the pressure. 47

That makes the transition difficult financially. Search advertising has historically provided a large, relatively established source of cash, whereas AI infrastructure and applications require substantial investment before they can produce comparable scale or margins. Baidu is therefore managing two opposing forces at once: protecting the long-term quality of an AI-first search product while accepting weaker near-term monetization. 245

AI growth is substantial, but not yet enough

Baidu Core AI-powered Business revenue reached RMB12.5 billion, up 25% year over year and equal to 50% of Baidu General Business revenue. The figure shows that AI is no longer a peripheral experiment in Baidu’s business mix. 35

The strongest growth came from infrastructure. AI Cloud Infra revenue rose 50% to RMB7.3 billion, while GPU-cloud revenue increased 283% year over year. Those results point to growing demand for the computing capacity needed by businesses adopting AI. 3517

But growth in one segment does not automatically offset decline in another. Baidu General Business revenue was RMB25.2 billion, down 4% year over year, underscoring that the expanding AI portfolio has not yet reversed the broader contraction. 6

Why the pivot is weighing on profit

Baidu is investing across foundation models, AI cloud, applications, proprietary chips and autonomous driving. The strategy is designed to connect the company’s search distribution and data with the infrastructure and technologies needed to deliver AI services at scale. 15

Those investments also explain why stronger AI revenue has not translated into a recovery in net income. Cloud capacity, model development, chip design and autonomous-driving systems involve upfront spending and execution risk. They may create larger, more recurring businesses over time, but their financial payoff is less immediate than advertising revenue. The available Q2 figures show commercial traction in AI cloud and core AI-powered operations; they do not yet show that these businesses have fully replaced the profitability of the legacy model. 135

Autonomous driving and chips extend the AI strategy

Autonomous driving and chip design are longer-horizon bets rather than immediate substitutes for search ads. Autonomous driving gives Baidu a way to deploy its AI capabilities in the physical world, while its Kunlunxin proprietary chips extend the strategy into the computing hardware that supports AI workloads. 1

This broadens Baidu’s ambition beyond adding generative features to search. The company is attempting to build an integrated platform spanning models, cloud infrastructure, applications, chips and mobility. The benefit would be more control over the technology stack and more potential routes to monetization; the trade-off is greater capital intensity and a longer path to returns. 13

What Q2 says about Baidu’s transformation

Baidu’s results do not show a failed AI strategy, but they do show an incomplete one. AI-powered revenue has reached half of General Business revenue, and cloud demand is growing rapidly. At the same time, online marketing revenue is falling by nearly one-fifth, total revenue is declining for a fifth consecutive quarter, and profit remains sharply below the prior year. 35913

The key question is no longer whether Baidu is investing in AI. It is whether AI cloud, applications, chips and autonomous driving can scale quickly—and profitably—enough to compensate for the advertising business that is being weakened during the transition. Q2 suggests the building blocks are becoming commercially meaningful, but the legacy-business decline is still outpacing the payoff from the new platform. 235