Guotai Haitong: Personal agents reshape consumer entry points, platform ecosystem value faces a revaluation window

Guotai Haitong released a research report stating that personal agents are rewriting consumer entry points and are expected to unlock a trillion-level consumer market. Overseas players Muse and Instinct are leading, while in the domestic market Tencent is best positioned, and WeChat Xiaowei is expected to drive a revaluation of platform ecosystem value. Guotai Haitong's main views are as follows: Personal agents are rewriting consumer entry points, with broad market space. As products such as Muse and Instinct rapidly gain popularity, consumer-grade AI is undergoing a paradigm shift from passive conversational chat Siasun Robot&Automation to proactive execution-oriented autonomous agents. The industry focus is shifting from "answering questions" to "completing tasks," and these agents are beginning to handle complex, high-value personal workflows such as travel booking, bill negotiation, schedule management, and shopping. As a new type of consumer entry point, personal AI agents are expected to unlock a trillion-level consumer market. According to calculations by PHD in collaboration with WARC, global consumer spending facilitated by Agentic AI will rise from approximately US$944 billion in 2026 (1.3% of private consumption) to US$3.35 trillion in 2030 (3.8%), with a CAGR of 37.3%. China's internet sector has a high degree of platform integration, which is conducive to mature platforms implementing consumer-grade personal agents. Muse and Instinct show standout data, leading the consumer-grade agent track. (1) Muse: It runs on an independent cloud virtual machine. Users can issue instructions through the Muse App, WhatsApp, or the web, and the Agent autonomously breaks down tasks, operates web pages and external applications, and supports continuous follow-up on long-cycle tasks. Deeply tied to the Meta ecosystem, it relies on the approximately 3.6 billion daily active user ecosystem base of the app family, with distribution capability and contextual data forming its core advantages. Within 1.5 months of launch, more than 10 million people in the US and Canada had activated it, with about 25% using it daily. Shopping is the largest use case, with more than 25% of weekly active users shopping via Muse, spending over US$200. It offers a free basic version plus two subscription tiers, with medium- to long-term focus on monetization through transaction commissions. (2) Instinct: It has no standalone app and uses a dedicated phone number, email, and cloud computer as its. Users can issue instructions by sending messages via iMessage or WhatsApp or by making a phone call, and the Agent directly executes tasks such as bookings, shopping, cancellations, and schedule coordination. As of September 2026, Instinct's user base exceeded 100,000, growing at a day-over-day rate of 10%-11%. The platform's annualized GMV has exceeded US$1 billion, about half of which comes from travel-related transactions. Shopping users spend an average of more than US$1,300 per person per month. About 40% of users link their primary credit card within three weeks of use, and the retention rate after linking sensitive assets is about 80%. It is currently completely free with no advertising, and the long-term business model concept is to take commissions from facilitated transactions. WeChat Xiaowei: Benchmarking the Chinese market, Tencent is the most advantaged player. Xiaowei has been in gray testing since June 2026, providing an embedded AI experience within WeChat. By connecting the mini-program ecosystem, it is expected to reshape its service distribution logic and may eventually achieve an Agent-to-Agent transaction closed loop in the long term. It adopts a multi-model routing mechanism combining WeLM and third-party models, focusing on user privacy protection, adaptation to native WeChat scenarios, and cost efficiency. Traffic entry points, scenario closed loops, and contextual data build its core barriers. It is currently expanding volume in stages, and the investment scale is expected to be smaller than Tencent's investment in Yuanbao over the past year. In the long term, it is expected to improve the overall monetization efficiency of the WeChat ecosystem and drive a revaluation of the platform's long-term ecosystem value by increasing users' WeChat usage frequency, expanding transaction scale and increasing payment frequency, enhancing advertising conversion, and promoting enterprise service demand. Risk warnings: commercialization falling short of expectations, AI investment affecting current-period profit, data security risks, policy risks, and macroeconomic volatility.