In a move underscoring its commitment to platform integrity, social media giant Xiaohongshu, often dubbed “Little Red Book,” unveiled its anti-fraud initiatives for the first half of 2025 at a recent governance open day held in Beijing. The company is aggressively tackling the proliferation of inauthentic content, aiming to cultivate a more trustworthy environment for its users and advertisers.
The report highlighted a significant clean-up effort, detailing a robust crackdown on misleading practices. Over the first six months of the year, Xiaohongshu took action against 3.2 million fraudulent posts, removed 10,000 accounts built on fabricated personas, and addressed 600,000 instances of low-quality, AI-generated content. Moreover, the platform banned over 10 million accounts associated with malicious, industry-scale fraud operations.
Xiaohongshu’s anti-fraud strategy centers on three primary areas of concern: false and misleading content, fabricated user profiles, and AI-generated content designed to deceive. These areas are viewed as key vulnerabilities for advertisers, intellectual property, and user trust.
To combat deceptive content, Xiaohongshu has deployed a suite of advanced technologies, including 15 sophisticated AI models designed for content identification. These models, which analyze content information, extract critical summaries, and automatically detect AI-generated rumors, have allowed the platform to intercept fraudulent material in real-time and conduct extensive retrospective sweeps. This has been instrumental in the platform’s ability to remove millions of fraudulent posts.
Furthermore, the platform has actively addressed fake user profiles by implementing a system of pre-content review and introducing several image and video recognition tools. This proactive approach has been effective in accurately identifying and removing accounts that are not genuine, including those attempting to impersonate real users to generate fraudulent engagement.
Recognizing the growing sophistication of artificially generated content, Xiaohongshu has invested heavily in improving the precision of its AI content detection. This is done to limit the reach of low-grade AI content and to increase transparency by proactively labeling AI-generated material. These efforts led to the removal of about 600,000 instances of deceptive, low-quality AI posts during the first half of 2025.
The platform has further enhanced user reporting mechanisms, providing several avenues for users to flag suspicious content, though details on these were not provided. Data from the first half of the year shows that the average processing time for suspected false content decreased by 25%, while the volume of user-generated reports increased by 40%. These figures underscore the increased awareness and engagement of platform users as a result of the company’s efforts.
Looking ahead, Xiaohongshu aims to refine its reporting process further, with plans to simplify user actions for flagging potentially misleading content. The goal is to let a user submit a single report through a single screen.
Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/2331.html
