As scrutiny of algorithmic practices intensifies in China, short video app Kuaishou announced plans to use AI-driven tools to combat rumors, improve platform governance, and amplify “positive” content.

Several platforms, including Douyin, the Chinese version of TikTok, lifestyle app Xiaohongshu, and shopping platform Pinduoduo, have recently said they intend to improve transparency regarding their algorithms and platform management.

These moves come amid rising concerns about online fraud targeting elders and minors, as well as growing complaints of privacy breaches and the manipulation of user habits.

According to Kuaishou, the platform has introduced tools to allow users greater control over content preferences, enabling them to customize recommendations and opt out of algorithmic tracking. To boost transparency, the platform plans to explain its algorithm principles in accessible terms and provide regular updates on system adjustments.

Focusing on content diversity, Kuaishou announced a “positive algorithm” initiative to promote trustworthy and high-quality material, vowing to invest 100 billion yuan ($13.6 billion) in traffic and 1 billion yuan in cash incentives planned for 2025. Additionally, the platform is enhancing efforts to combat misinformation by optimizing algorithms to prioritize fact-checking and rumor-debunking content.

Kuaishou has also rolled out features tailored for vulnerable users, including an “Elderly Mode” with simplified interfaces, and a one-click “Minors’ Mode” developed in partnership with mobile manufacturers.

These measures come after a November warning from public security authorities over regulatory lapses, including inadequate implementation of the “Minors’ Mode,” which led to penalties and a mandated review of platform practices.

The government’s focus on short video platforms is tied to their massive influence. As of September 2024, mobile internet users in China totaled 1.25 billion — over 88% of its population — with short videos accounting for nearly 23% of daily online activity, second only to messaging apps.

This dominance has raised concerns about the platforms’ role in fostering content homogenization, manipulating trending topics, and engaging in algorithm-driven pricing discrimination.

To address these issues, authorities have intensified regulatory scrutiny, mandating platforms to conduct self-checks, enhance algorithm security, and promote content diversity.

In late December, over 100 platform executives were given legal training on these reforms, signaling a systemic push for accountability. While many users have welcomed these measures, some remain skeptical, criticizing vague disclosures and the limited transparency of algorithmic systems.

“Algorithms often operate as proprietary secrets, making transparency a challenge,” noted Cao Wei, a commercial law professor at Southwest University of Political Science and Law, in the southwestern metropolis of Chongqing. He emphasized the need for balancing public oversight with the protection of intellectual property.

(Header image: VCG)