When Youth Enter the Algorithmic Wild: Discovering and Understanding Potentially Harmful Teen Videos on Douyin and Kwai

2026-05-22

AI Summary

This study addresses the unclear scale and dissemination mechanisms of potentially harmful content recommended to adolescents on short-video platforms. The authors propose PHTV-Scout, a framework integrating survey data from 683 adolescents with a behavioral simulation system to capture authentic recommendation streams. Leveraging a LoRA-finetuned multimodal classifier, they identify and fine-grained categorize harmful videos. This work presents the first large-scale, behavior-driven measurement of adolescent exposure to harmful content, analyzing 186,000 videos and 51,000 comments. Findings reveal that 6.11% of recommended videos contain potentially harmful material, with 53.2% involving child sexual exploitation imagery. The study further uncovers how such content spreads through covert interactions and evasion tactics, demonstrating that Youth Mode can fully block these recommendations—yet its adoption remains low at only 30–41%, highlighting a critical protection gap.

Abstract

Short-video platforms like Douyin and Kwai have become central to adolescent digital life, but they also risk exposing teens to algorithmically amplified harmful content. Despite its societal importance, the scale, mechanisms, and real-world impact of this exposure remain poorly understood. Measuring it is challenging: recommendation feeds are personalized black boxes, harmful content employs sophisticated evasion tactics, and naive crawlers fail to replicate authentic teen behavior. To bridge this gap, we propose PHTV-Scout, the first large-scale, behaviorally grounded measurement framework for Potentially Harmful Teen Videos (PHTVs). We integrate an offline survey of 683 adolescents with a tri-module online pipeline: (1) PHTV Hunter simulates teen accounts to collect recommendation feeds; (2) PHTV Arbiter, a LoRA-finetuned multimodal classifier, detects PHTVs with 94.29% accuracy and 96.41% precision; and (3) PHTV Analyzer performs fine-grained categorization and impact assessment. Over six months, we analyzed 186,727 videos and 51,287 comments, uncovering a troubling 6.11% PHTV prevalence--dominated by Child Sexual Exploitation Imagery (53.2%)--and revealing that harmful content thrives through covert interactions (e.g., grooming comments, self-disclosure) and active evasion (semantic camouflage, noise injection). Crucially, while Youth Mode blocks 100% of PHTVs, its low adoption (30-41%) leaves most teens unprotected. We further show that exposure is driven not by user identity but by regulation, platform algorithms, and even passive browsing, exposing the fragility of adolescent information environments. Our findings call for a paradigm shift from reactive takedowns to proactive, human-centered safeguards.

Problem

• Potentially Harmful Teen Videos
• algorithmic recommendation
• adolescent exposure
• content moderation
• short-video platforms

Innovation

• PHTV-Scout
• LoRA-finetuned multimodal classifier
• behaviorally grounded measurement
• algorithmic amplification
• semantic camouflage