What type of study is this?

This is a Quantitative Study study.

October 12, 2025Open Access

Towards an Automated Framework to Audit Youth Safety on TikTok

Key Points

Preliminary findings indicate minimal differences in harmful content exposure between youth and adult accounts, raising safety concerns.
Over 7000 videos were analyzed to classify harmful versus non-harmful content, emphasizing youth exposure risks.
Simulation of interactions via sockpuppet accounts showed challenges in detection accuracy using both large language and vision-language models.
The research identifies the need for improved transparency in TikTok's content moderation strategies and enhanced youth safety measures.

Abstract

This paper investigates the effectiveness of TikTok's enforcement mechanisms for limiting the exposure of harmful content to youth accounts. We collect over 7000 videos, classify them as harmful vs not-harmful, and then simulate interactions using age-specific sockpuppet accounts through both passive and active engagement strategies. We also evaluate the performance of large language (LLMs) and vision-language models (VLMs) in detecting harmful content, identifying key challenges in precision and scalability. Preliminary results show minimal differences in content exposure between adult and youth accounts, raising concerns about the platform's age-based moderation. These findings suggest that the platform needs to strengthen youth safety measures and improve transparency in content moderation.

Towards an Automated Framework to Audit Youth Safety on TikTok

Key Points

Abstract

Cite This Study

Also Consider

Also Consider