WeChat Refutes Privacy Rumors Surrounding Its Native AI Assistant 'Xiao Wei'

A WeChat employee publicly clarified how the platform's native AI assistant 'Xiao Wei' handles user data, refuting claims that it secretly reads chat logs or invades privacy. The assistant only accesses client-side chat history temporarily when explicitly invoked by the user, immediately discarding raw logs after generating summaries. Privacy concerns are a major hurdle for integrating AI assistants into core messaging platforms that handle sensitive personal communications. By detailing a strict 'use-and-discard' local processing model, WeChat aims to build user trust while testing broad AI features like automated summarization and agent-to-agent interaction. Xiao Wei operates strictly within the user's existing permission scope, meaning it cannot view deleted Moments or unauthorized content. Furthermore, chat logs used for summarization are processed temporarily on the local client without being saved on servers.