In-Depth: Alibaba, Tencent’s Expansion Into AI Creates Opportunities (AI Translation)

【周刊提前读】深度：阿里腾讯全线布局AI 带来什么机会？

Alibaba and Tencent, China’s two largest internet companies, are actively repositioning themselves to capitalize on the AI boom, mirroring global tech giants seeking to benefit from advancing AI models and applications. After organizational revamps and strategic shifts, both companies have entered a new investment phase, focusing especially on AI infrastructure and applications[para. 1]. In February 2024, DeepSeek, a rapidly emerging Chinese AI model company, gained intense industry attention, its founder Liang Wenfeng sharing the stage with Alibaba and Tencent founders at a high-profile entrepreneurs’ meeting[para. 2]. This set the stage for Tencent’s “Yuanbao” AI chat app and Alibaba’s massive three-year investment plan of over 380 billion yuan in cloud and AI hardware, exceeding the company’s total capital expenditure over the past decade and setting a record for Chinese private enterprise investment in this sphere[para. 4].

The DeepSeek breakout showcased how high-quality foundational models could be built with limited resources, disrupting the notion that only vast compute and traffic pools matter in China’s AI race. Tencent quickly integrated DeepSeek’s model into Yuanbao, accelerating its update cycle—36 updates in four months, helping its daily active users (DAU) skyrocket by 20 times in March and placing it among China’s top three AI chat apps, alongside ByteDance’s Doubao and DeepSeek itself[para. 5][para. 7]. Meanwhile, Alibaba leveraged DeepSeek’s popularity to highlight the flexibility of its open-source Qwen AI models, which now offer over 100,000 derivatives, catering to various developer needs and lowering entry costs compared to rivals like DeepSeek R1[para. 8][para. 10].

Strategically, Alibaba and Tencent have adopted different priorities: Alibaba continues to strengthen core AI models and “AI+Cloud” services, doubling down on the Model-as-a-Service (MaaS) sales model, while Tencent focuses on rapidly embedding AI functionality in consumer applications, especially leveraging vast user bases from products like WeChat, which now supports AI chat and image analysis through Yuanbao[para. 12][para. 16][para. 22][para. 27]. Both groups, however, maintain all-encompassing AI portfolios, investing in basic model platforms, industry applications, and cloud infrastructure. For example, Alibaba rolled out the latest Qwen3 model in March, claiming superior performance and considerably lower deployment costs, while Tencent persistently updates its foundational “Hunyuan” model and related AI agents[para. 13][para. 15][para. 21].

In addition to internal development, Alibaba and Tencent have become the most active AI strategic investors in China, backing top AI startups like Zhipu AI, MiniMax, Baichuan Intelligence, and Moonshot AI, among others[para. 46][para. 48]. Both are also grappling with the high demand for computing power, prompted by US chip export restrictions and the need to diversify hardware sources. As of 2024, Alibaba led with over 23,000 Nvidia AI chips and piloted large-scale investments in domestic chip alternatives, while Tencent and ByteDance also ramped up their procured and self-developed chips to ensure AI development continuity despite supply constraints[para. 147][para. 149].

The companies’ internal operations and main businesses, such as Alibaba’s B2B platform and Tencent’s gaming and advertising divisions, are undergoing deep AI-driven transformations. Alibaba’s international B2B search engine Accio, for instance, uses AI to turn vague buyer requests into actionable business proposals, significantly improving transaction efficiency. Taobao, Quark browser, and DingTalk are also embedding AI features for consumer and enterprise users[para. 102][para. 106][para. 109]. Tencent’s upgrades have helped applications like QQ Browser and WeiXin/CWeChat offer AI search, knowledge management, and personal productivity solutions, while its advertising business harnesses AI to improve targeting and engagement rates, with some ad placements’ click-through rates increasing to 3%[para. 118][para. 120][para. 121].

In sum, while Alibaba bets heavily on “open source + cloud” for AI proliferation, and Tencent leverages its massive application ecosystem and content resources, both are poised—alongside ByteDance—to become China’s dominant AI champions. The precise business models for monetizing core AI models remain in flux, but both giants are racing to create compelling user-facing AI experiences, secure computational resources, invest in promising startups, and weave AI into every facet of their sprawling digital empires, setting new standards for China’s AI industry[para. 176][para. 181].