China's AI sector has pivoted from a strategy of chasing technological parity to one of aggressive value capture, with a focus on applications and monetization that is already reshaping the investment landscape, according to a new flagship report from Morgan Stanley. The bank's research indicates the market is underpricing the velocity of this change, as Chinese AI adopters have seen their forward 12-month earnings per share estimates jump approximately 62% since the end of 2023, dramatically outpacing the broader MSCI China index.
"The story of AI in China is no longer about catching up, but about rewriting the rules of the game," the Morgan Stanley report states. The bank argues that while Western firms focus on frontier model breakthroughs, China's edge lies in deployment speed, cost efficiency, and system-level integration, creating a distinct competitive advantage in turning AI potential into tangible financial results.
This strategic shift is backed by improving financials. The report projects that the EBIT profit margin for Chinese companies adopting AI is set to expand from about 4% in 2021 to a range of 16% to 17% by 2027. This margin expansion, driven primarily by cost efficiencies rather than revenue growth, underscores the immediate bottom-line impact of AI deployment. A Morgan Stanley survey of Chinese CIOs reinforces this trend, with 47% planning to initiate their first AI projects in the coming year, up from 40% in the previous survey.
The core of China's AI advantage, according to the report, is delivering comparable model intelligence at a fraction of the cost. By leveraging architectural innovations like Mixture-of-Experts (MoE), model distillation, and hardware-software co-optimization, Chinese firms can achieve inference costs that are just 15% to 20% of their U.S. counterparts. This efficiency has allowed the market share of top Chinese large language models (LLMs) in token usage to surge from 5% in April 2025 to 32% by March 2026, according to OpenRouter data.
From Price Wars to Pricing Power
The competitive dynamics in China's foundational model layer have undergone a significant reversal. After a period of intense price wars in 2024 that saw API costs slashed by 70% to 90%, leading providers like Alibaba, Baidu, and Tencent, alongside startups like Z.ai and MiniMax, are now implementing price hikes for their new flagship models. Morgan Stanley data shows average API input prices climbed by about 80% between the second quarter of 2025 and the first quarter of 2027.
This is not a cost-driven increase but a reflection of performance gains, signaling a market shift from commoditization to value-based pricing. Z.ai, for instance, increased prices by over 200% for its GLM-5 model compared to its predecessor. Morgan Stanley identifies Alibaba as the best-positioned full-stack AI platform, with its integrated portfolio spanning the T-Head chip division, Alibaba Cloud, the Qwen model series, and consumer applications.
The Semiconductor Self-Sufficiency Push
Underpinning this application boom is a concerted drive toward semiconductor independence. Morgan Stanley projects the domestic supply rate for AI chips in China will climb from 41% in 2025 to 86% by 2030, with the total market size expanding from approximately $19 billion to $67 billion in the same period.
The focus of procurement is shifting from peak performance to "deployable cost efficiency." Channel checks by the bank indicate that domestic AI accelerators can offer a 30% to 60% lower total cost of ownership (TCO) compared to currently available Nvidia products in China, with comparable per-token inference costs. This is creating a robust internal market for domestic champions as export controls tighten.
New Frontiers: Power, Robots, and Autonomy
The report also highlights emerging growth vectors where China's manufacturing base and data advantages create unique opportunities. The explosive growth in AI data center demand is shifting the bottleneck from compute to power, creating a massive new market for energy storage systems (ESS). Morgan Stanley forecasts that annual data center ESS deployments in China will reach 85 GWh by 2030.
In robotics, the market is approaching a scale inflection point. The bank estimates Chinese sales of humanoid robots will grow from 12,000 units in 2025 to over one million by 2034, tapping into a potential $1 trillion domestic market by 2050. Similarly, the automotive sector is nearing an AI-driven turning point, with regulators issuing the first L3 autonomous driving permits. The penetration of L2+ intelligent driving systems is expected to surpass 50% by 2030, up from about 25% in 2025.
This article is for informational purposes only and does not constitute investment advice.
