Why Zhihu Forced Login and What It Means for Users and AI Training

In early 2022, China’s Ministry of Industry and Information Technology issued a directive to stop apps from forcing users to download them in order to read full articles or comments, aiming to improve web browsing experience. Major sites such as Baidu, Zhihu, and Sina quickly removed the mandatory app‑download prompts.

Despite this, many websites continued to restrict full content to logged‑in users. Zhihu, for example, began in May of the previous year to block non‑logged‑in visitors from reading entire answers on its web platform; clicking “expand full text” triggered an automatic login popup.

Zhihu also altered its robots.txt to hinder search engine crawlers and, according to user‑inspected page source, does not transmit the complete answer data to unauthenticated requests, making it impossible to view full answers even with technical workarounds.

Historically, Zhihu has been unfriendly to anonymous users. Since 2020 the site required login to access any content, and the login wall has contributed to user dissatisfaction, reduced search engine visibility, and even app uninstallations.

Industry observers speculate that Zhihu’s strict login policy serves a defensive purpose: preventing its content from being scraped by AI companies for training large language models.

Performance data shows a decline in user metrics: in Q3 2024 Zhihu reported an average monthly active user count of 81.1 million and 16.5 million paid members, down from 101.2 million MAU and 5.5 million paid members in Q3 2021.

Possibly in response to this user loss, Zhihu recently removed the “login to view full text” restriction. According to recent user feedback, PC users can now click “expand full text” and read the entire answer without encountering a login prompt.

This change improves accessibility for many users, but the broader implications for content protection and AI data usage remain a topic of discussion.