Abstract: In 2025, China's AI search landscape is fiercely competitive, with traditional giant Baidu and Alibaba's rising star Kuake locked in a battle for supremacy.

For Baidu and Alibaba, the apparent AI-to-consumer focus belies distinct underlying motivations.

Baidu grapples with waning traffic from short video platforms and evolving AI dialogue interactions, which disrupt its traditional revenue streams. Conversely, Alibaba faces shrinking profit margins due to a slowing e-commerce market and intensified competition, necessitating a new growth trajectory.

Both companies share a common goal: developing one-stop solutions.

Below is the main text:

The Clash Between Traditional Search Giants and AI Newcomers

As AI products proliferate, a silent competition in search technology intensifies.

Baidu's search interface now features an AI voice assistant that proactively answers queries and offers solutions. Xiaohongshu's image and text notes break the traditional browsing mold, enabling users to trigger intelligent Q&A with a simple tap. Even everyday input methods now embed search functions within the input box.

The AI search revolution ignited by ChatGPT in 2023 has reached a fever pitch this year.

Five years ago, the search engine landscape was dominated by Baidu, Google, and Bing. Today, it includes established players like Google, Baidu, and 360, alongside newcomers with cutting-edge AI technology such as OpenAI and DeepSeek. Content platforms like Zhihu, Xiaohongshu, and Bilibili have also entered the fray, with internet giants like Tencent, ByteDance, and Alibaba joining with their respective ecosystems. Traditional search engines, once entrenched, are now adapting to their "mid-life crisis".

Take Baidu, for instance; it no longer suffices with merely listing links. The 2025 launch of Baidu AI assistant marks a significant step in its transformation.

For example, when a user searches for "how to prepare for IELTS," they receive a detailed phased study plan, complete with analysis of common errors, rather than a list of web links.

On the other hand, Kuake disrupts the status quo with its Deep Search functionality. Addressing issues like AI content hallucinations and irrelevant answers, Deep Search has garnered positive user feedback.

Unlike most deep-thinking AI that follows the "search first, think later" logic, Kuake's Deep Search adopts a unique "think first, search later" model. This innovation in underlying logic has propelled a leap in AI search's intelligent retrieval capabilities.

Highly Synchronized Development Trajectories

Baidu Wenku and Kuake serve as the "vanguards" for Baidu and Alibaba, respectively.

Both internet giants build upon Chatbot technology, deeply integrating intelligent dialogue, deep retrieval capabilities, and diverse AI tools, aiming to create a service closed loop for "one-stop task resolution".

Starting from search engine core technology, Baidu Wenku (now integrated with Baidu Netdisk technology) has effectively transformed into a comprehensive tool platform.

Kuake attracts young users through vertical domain applications like Kuake Netdisk and Kuake College Entrance Examination, ultimately constructing an "AI Super Frame" technology integration architecture.

Over a decade ago, Li Yanhong proposed the "Frame Computation" concept, enabling users to complete information acquisition and task processing within a unified technical framework without jumping between heterogeneous systems. This resonates with Kuake's current technical architecture.

Secondly, from the perspective of internet giants' ecological technology layout, precipitating user traffic and data assets within their technical systems is a common core demand.

Under the backdrop of large model and cloud computing technology iterations, vertical tool integration and one-stop solutions emerge as key breakthrough points for both Baidu and Alibaba.

Du Yu, a former technology investor at Sequoia Capital, noted that students and professionals are the core user groups for AI applications, with desktop office work as the primary usage scenario. This overlaps significantly with Kuake and Baidu Wenku's existing user bases, providing a natural testing ground for technological innovation.

Commercially, Baidu Wenku boasts a mature document payment ecosystem. With AI function empowerment, it has forged new commercial growth points. Data shows over 90 million monthly active users for AI functions, with over 40 million paid users, demonstrating market acceptance of technical value-added services.

Kuake forms a closed loop of user value through value-added services like cloud storage and intelligent scanning. With Alibaba Group's technical support, it prioritizes user scale growth and smart feature penetration rate enhancement.

Differentiated Business Models

While Baidu and Alibaba focus on AI-to-consumer applications, their underlying motivations differ.

Baidu contends with waning traffic dividends and a single income structure due to short video platforms' rise and AI dialogue interaction innovations. Conversely, Alibaba grapples with compressed profit margins amid slowing e-commerce growth and intensified competition, necessitating a breakthrough from traditional retail and a new value growth curve.

Alibaba chooses to enhance its C-end application portfolio through Kuake.

Alibaba Cloud's infrastructure, Tongyi Qianwen large model's algorithmic support, and DingTalk's enterprise services form a technical closed loop with Kuake's C-end traffic.

Kuake serves as a user verification window for the Tongyi large model, extending its technological influence from the developer community to the consumer market.

This synergy is evident in hardware, with Tmall Genie and Kuake's integration and the upcoming release of AI smart glasses at year's end, indicating Alibaba's push into terminal entry points via software+hardware bundling.

Baidu, on the other hand, selects AI Agent as its differentiated path. Tools like "Free Canvas" and "Miaoda" focus on vertical scenarios like PPT generation and code development, activating paid potential through deep technical empowerment.

Miaoda data shows an equal split between enterprise and individual users among beta testers, reflecting Baidu's strategy to convert technical depth into commercial value through bidirectional To B and To C scenario penetration.

Unlike Alibaba's hardware ecological layout, Baidu emphasizes niche area technical moats, aiming for user paid conversion by solving specific task efficiency pain points.

Internal business synergy is crucial in this competition.

Alibaba has merged Tongyi APP into Kuake's Intelligent Information Business Group, completing the technological link from large model R&D to C-end applications. Baidu Wenku accesses external models like DeepSeek, reflecting a balanced strategy between self-developed technology and ecological cooperation.

Companies like ByteDance and Tencent are also accelerating their "super assistant" layouts, with industry competition evolving from single product function iterations to a scenarios+technology+ecology three-dimensional game.

Epilogue

This year, the global AI search technology regulatory framework is rapidly forming.

The European Union mandates clear labeling of generated content's data sources, while China has introduced "search result credibility ratings" standards, suggesting future market competition will shift from algorithmic performance and ecosystem construction to technical compliance and user trust systems.

This technological progress necessitates users' adaptation to new information interaction modes—from passive web link receipt to active intelligent system conversation, and from blind information retrieval to resource integration. The human-computer interaction form is undergoing a fundamental transformation.

Anticipating user needs is crucial for creating a successful search system product.

- XINLIU -