WeChat and Alipay Counterattack Against Doubao: Turning Mini-Programs Into AI Skills
In a coordinated move that signals a major strategic shift, China's two dominant super-apps — WeChat and Alipay — have both moved to transform their vast ecosystems of mini-programs into AI-callable "Skills," directly countering the rise of ByteDance's AI assistant Doubao. On June 2, the Financial Times reported that WeChat was internally testing an AI Agent capable of invoking mini-programs to complete user tasks. Days later, on June 8, WeChat publicly released technical spe
In a coordinated move that signals a major strategic shift, China's two dominant super-apps — WeChat and Alipay — have both moved to transform their vast ecosystems of mini-programs into AI-callable "Skills," directly countering the rise of ByteDance's AI assistant Doubao. On June 2, the Financial Times reported that WeChat was internally testing an AI Agent capable of invoking mini-programs to complete user tasks. Days later, on June 8, WeChat publicly released technical specifications for connecting mini-programs to AI, with JD.com, Meituan, and Didi immediately announcing integration. Then on June 15, Ant Group signaled the launch of an "AI-powered Alipay" that allows users to switch to a fully conversation-driven interface capable of tapping into millions of mini-programs. Both platforms are following a similar playbook: they are encouraging merchants to proactively convert their services into MCP/Skill formats that AI can directly invoke, while also enabling AI to interact with unmodified mini-programs through screen-reading technology. The underlying logic is a fundamental shift from what the industry calls "distribution assets" to "invocation assets." For the past decade, mini-programs were valuable because users could discover them through platform entry points — search bars, service menus, and recommendation feeds. The new paradigm flips this model: mini-programs become valuable because AI can call them directly as tools. When a user says "order me a not-too-sweet latte," the AI automatically selects the right mini-program, fills in the parameters, and completes the order. The user no longer "sees" the mini-program; the AI "invokes" it. The implications are profound. Previously, traffic distribution was controlled by platform operations and user choices. In the new AI-driven model, traffic allocation shifts to the model's intent understanding — which service it selects and how it orchestrates multiple capabilities. The entry point is no longer a fixed door that users must pass through, but one of many routes the model can choose. WeChat's approach is notably cautious. Its AI entry lives in the widget screen (负一屏), and the company is being careful not to disrupt its established social ecosystem. Alipay, unburdened by social baggage, has taken a more aggressive approach with its one-click AI mode switch. The urgency behind these moves is Doubao. ByteDance's AI assistant has rapidly grown to over 300 million monthly active users by applying consumer app methodology — educational tutoring, voice companionship, and content creation. Doubao represents the "pure AI-native entry point" threat: if users get accustomed to accessing services through a completely new AI interface, WeChat and Alipay risk being bypassed entirely. The real battle is not about whether Doubao can write Skills — that's a technical problem anyone can solve. It's about time: whoever can convert their existing service modules into an AI-callable layer before users form new habits wins. WeChat and Alipay have a decade of accumulated mini-program infrastructure. Their Skill-ification is essentially a head start — wrapping existing services in AI interfaces before the pure-AI challengers can build their own service networks from scratch.
📌 Kaynak
Bu haber XML kaynağından derlenmiştir. Tamamı için orijinal habere gidin.
Orijinal haberi oku →