According to a report from video creation platform Kapwing, nearly 60% of videos shown to a brand-new TikTok account’s For You Page could be classified as “AI slop,” a term commonly used to describe low-quality AI-generated content designed primarily to attract views and engagement. The study found that 294 of the first 500 videos recommended to a new TikTok account fell into that category.
Kapwing’s researchers analyzed more than 10,000 TikTok videos across 20 categories and hashtags, while also conducting tests using newly created accounts to examine what content the platform recommends to first-time users.
The findings were later highlighted by Tubefilter, which reported that TikTok appears to recommend substantially more AI-generated content than competing platforms.
Children’s content showed some of the highest rates
The report found particularly high concentrations of AI-generated content in videos aimed at children. According to Kapwing, 57% of videos reviewed in TikTok’s Kids category were classified as AI slop. Some child-focused hashtags showed even higher levels, including #cartoonkids, where 97 out of 100 featured videos were identified as AI-generated.
Other categories with relatively high levels of AI-generated content included science, education, health and history, where AI narration and generated visuals are increasingly common. By comparison, categories centered around real people and physical demonstrations, such as fashion, music and fitness, showed much lower rates.
TikTok and YouTube show different results
Kapwing also compared TikTok with YouTube by conducting the same test on YouTube Shorts. The company found that 21% of the first 500 Shorts recommended to a new YouTube account qualified as AI slop, compared to 59% on TikTok.
The comparison comes as YouTube continues investing in AI-content moderation tools. Earlier this year, YouTube expanded its AI labeling systems, introduced automatic detection for synthetic content and continued rolling out deepfake detection tools designed to identify unauthorized AI-generated likenesses.
As AI video generation tools become more accessible, platforms are facing growing questions about content quality, transparency and how recommendation systems should balance AI-generated media with human-created content. The findings suggest those discussions are likely to continue as synthetic content becomes more common across social media platforms.