Morgan Stanley believes the market is undergoing a structural shift: despite the global AI race's clamor, China is carving out a unique path—an "open" model strategy pitted against the world's "closed" systems.

According to the report, as of January 8th, Morgan Stanley's latest research indicates that China now accounts for half of the world's top 10 SOTA (State-of-the-Art) models. The firm forecasts that China's total addressable market (TAM) for cloud AI will reach $50 billion by 2027. This suggests that, despite a complex external environment, the resilience of China's local computing power supply chain is strengthening, providing foundational support for an explosion in upper-layer applications.

For investors, the focus should shift beyond the infrastructure arms race towards the monetization capabilities and ecosystem barriers at the application layer.

The report explicitly states that China possesses significant comparative advantages in data volume, power supply, and its engineer dividend. Regarding specific investment picks, Morgan Stanley views Alibaba, with its combination of cloud computing and model capabilities, as the "best enabler" for China's AI development; meanwhile, Tencent, leveraging its WeChat ecosystem, holds the highest consumer-end (2C) monetization potential and high return on investment (ROI). This judgment implies that the pricing logic in capital markets will evolve from pure computing power speculation towards platform giants with massive user bases, proprietary data accumulation, and the ability to substantively deploy AI applications.

The application layer in China presents a unique landscape of "super-app" evolution progressing in parallel with an explosion of "AI-native applications."

Morgan Stanley particularly emphasized the immense potential of WeChat as a pioneer for AI Agents. As of July 2025, WeChat boasts 1.1 billion monthly active users (MAU), with a staggering average daily user time of 99.4 minutes and an average of 44.6 daily sessions per user. This high-frequency, deep user engagement provides fertile ground for the integration of AI Agents, enabling seamless integration into diverse scenarios like daily life, shopping, and travel.

Concurrently, AI-native applications such as ByteDance's Doubao, Baidu's ERNIE Bot, and Alibaba's Quark and Yuanbao are rapidly competing for user time. Data shows these applications are evolving from mere chatbots into emotional interaction tools, content creation aids, and even all-around AI assistants. For investors, this means companies with high user retention and rich contextual data will be the first to reap the traffic dividends brought by AI.

Beyond the buzz on the consumer side, AI penetration on the enterprise (2B) side is quietly reshaping industry landscapes.

Morgan Stanley's second-half 2025 China CIO survey reveals strong corporate willingness to deploy Generative AI (GenAI), shifting from early experimentation to substantive productivity enhancements. In vertical sectors like advertising, healthcare, finance, and energy, AI application scenarios are rapidly materializing. For instance, in advertising, AI is reshaping ad placement efficiency and content generation; in healthcare, AI image analysis and new drug研发 have become key drivers.

The survey also indicated that a significant proportion of work hours will be replaced by GenAI within the next three years, forecasting a shift in corporate IT spending towards AI-related infrastructure. In the SaaS sector, companies like Beisen are reshaping the human capital management market through features like AI interviewers and employee assistants; Kingsoft Office, via WPS AI, has not only increased user payment rates but also solidified its position in the office software market.

In its analysis of specific investment targets, Morgan Stanley provided precise, differentiated positioning for the major internet giants.

Alibaba is viewed as the "Best AI Enabler in China." The core logic is that Alibaba not only possesses the powerful Tongyi Qianwen (Qwen) large model but, more crucially, the deep integration of its Alibaba Cloud infrastructure with business scenarios like e-commerce and DingTalk allows it to capture value across the entire chain, from computing power and foundational models to the application layer.

Tencent is ascribed the evaluation of having the "Highest 2C Monetization Potential." Morgan Stanley pointed out that Tencent's AI strategy, integrated through the WeChat ecosystem, can achieve commercialization with an extremely high return on invested capital (ROIC). Its general AI assistant, Yuanbao, deeply integrates with WeChat Official Account content and search capabilities, giving it a natural barrier within the content ecosystem.

In contrast, ByteDance is defined as a "Full-Stack AI Leader," with the most comprehensive layout spanning from the underlying Volcano Engine, the middle model layer (Doubao, Seedance), to upper-layer AI-native applications (Jimeng, Kouzi, etc.) and even AI hardware. While Baidu started its AI transformation earliest, Morgan Stanley noted that its core advertising business is facing pressure from the transition to AI-powered search.

Therefore, investors should closely monitor the pace of AI implementation by these giants in their respective areas of strength, particularly focusing on those "pragmatists" capable of converting AI technology into tangible revenue and profit.