中国消费级 AI“激活年”，AI 应用胜负未定，摩根大通更看好 “二阶受益者”

JP Morgan pointed out that 2026 will be the "activation year" for consumer-grade AI in China, as chatbots transition from the trial phase to the user habit formation phase.

According to the Chasing Wind Trading Desk, on January 30, JP Morgan released a latest report stating that although large internet platforms are trying to shape the next traffic portal, the market is currently not ready to uniformly price the ultimate "winners and losers" of AI applications; early market share shifts reflect distribution capabilities rather than moats.

Based on this judgment, JP Morgan advises investors to avoid making black-and-white application layer bets at this stage and instead anchor trades on "second-order beneficiaries." This includes AI infrastructure benefiting from increased reasoning workloads, advertising platforms benefiting from heightened industry marketing intensity, and productivity enhancers in content-intensive industries.

JP Morgan specifically mentioned that Doubao, a subsidiary of ByteDance, has reached approximately 100 million daily active users (DAU), which is a milestone indicating that general chat interactions have crossed the threshold of the mass market. Meanwhile, according to Reuters, OpenAI's ChatGPT has reached 800 million weekly active users, further confirming the global trend that consumer-grade AI will experience rapid compound growth once product practicality and distribution align.

As chatbots evolve into high-frequency traffic portals, JP Morgan expects fierce competition among major internet platforms in China by 2026. This competition will have two direct market impacts: first, conversations will shift from occasional queries to daily multi-turn interactions, accelerating Token consumption; second, platforms will increase customer acquisition spending to compete for mind share, directing budgets towards performance-oriented channels.

Infrastructure and Advertising: The Most Certain "Second-Order Exposure"

JP Morgan emphasized in the report that it is still too early to trade on the final landscape of the chatbot market in China. Short-term share fluctuations mainly reflect distribution routes, marketing rhythms, and product iteration speeds, and have not yet formed solid competitive barriers. Therefore, the bank believes that a more feasible investment logic is to target the certain beneficiaries in the industry's advancement of AI interaction processes.

In terms of infrastructure, the growth of daily multi-turn chatbot conversations will directly benefit reasoning demand and Token throughput. As agents expand from simple conversations to deep queries and executions, Token consumption is expected to achieve compound growth, thereby increasing the utilization of cloud and AI technology stacks. JP Morgan believes that Alibaba and Baidu can clearly reflect the Token growth theme, and their value chains will reward the integration of scaled infrastructure and ecosystem levels.

In advertising and marketing, as giants like Alibaba and Tencent compete to shape new traffic portals, industry sales and marketing intensity will significantly increase. JP Morgan pointed out that spending on customer acquisition and repeated interactions across platforms will rise, while performance-oriented channels will capture a larger share of budgets. This creates a favorable trading backdrop for Kuaishou Technology, allowing it to benefit from the marketing dividends brought by industry competition even before the ultimate "winners" are determined

Enterprise Adoption: ROI Realization in High-Stakes Workflows

The report points out that the turning point for enterprise AI adoption lies in the measurability of value propositions in high-stakes workflows. Software development has become the clearest validation scenario. Alphabet CEO Sundar Pichai has stated that over a quarter of Google's new code is generated by AI; Tencent has also disclosed that its AI coding tool CodeBuddy reduces coding time by 40%.

JP Morgan believes that this data validates the underlying mechanism that AI assistants can translate model capabilities into throughput improvements for task-critical workflows. In addition to coding, similar trends are also evident in areas such as customer operations. According to Reuters, Klarna has revealed that the work efficiency of its AI assistant is equivalent to that of 700 employees. In China, platforms are productizing AI into enterprise toolkits, which will benefit foundational model developers and platform AI suppliers in both financial (capturing workload demands) and operational (reducing R&D costs) aspects.

Agent Commerce: From "Dialogue" to "Execution"

Following consumer adoption of chatbots, JP Morgan believes the next step is "agent commerce," where interfaces not only answer questions but also initiate and complete transactions. Alibaba recently upgraded its Qianwen application, integrating services like Taobao and Alipay, allowing users to order takeout and book travel within the chat interface, marking the realization of this transformation in China.

The report analyzes that the short-term opportunity for this transformation lies in improving funnel efficiency rather than an immediate disruption of market share. Conversational interfaces can reduce friction between discovery, comparison, and checkout, thereby improving conversion and retention within the ecosystem. Although various verticals (such as OTA) face long-term risks of upstream entry interception, given the complexity of services like travel, JP Morgan expects the market to undergo a sustained experimentation period before structural share changes become clear.

Gaming Industry: AI Accelerating Supply and Genre Innovation

Regarding the gaming industry, JP Morgan believes that the Chinese gaming sector will maintain strong performance in 2026, and AI adoption is an underestimated boon. The main impact of AI is not in replacing creativity but in compressing the "idea to market" cycle, accelerating asset creation and iteration. According to Reuters, Tencent has released an open-source 3D generation model, indicating that leading publishers are actively integrating AI as a productivity tool into their pipelines.

Moreover, the stability of the regulatory environment also supports the industry. Data shows that in 2025, China's National Press and Publication Administration approved 1,771 games, setting a new high since 2018, indicating a healthier content reserve. In terms of game genres, the popularity of "search and destroy" games like "PUBG" derivatives and "Dark Zone Breakout: Infinite" is also seen as a positive signal for industry growth