TikTok Shows 3x More AI Slop Than YouTube, According to Kapwing Report
Key Takeaways
- ▸59% of TikTok videos shown to new users are AI-generated—nearly 3x the rate on YouTube (21%)
- ▸Children's content is disproportionately affected, with some hashtags reaching 97% AI-generated content
- ▸Categories relying on visual illustrations and voiceovers are most affected; those requiring on-camera presence have minimal AI content
Summary
A comprehensive report from Kapwing, a video creation platform, found that 59% of videos served to new TikTok users in their For You feed are AI-generated content ("AI slop"), nearly three times the rate found on YouTube's Shorts. Kapwing's research involved manually reviewing over 10,000 TikTok videos across 20 popular categories, plus a separate fresh-account test where researchers scrolled through 500 For You videos. By comparison, the same methodology found that only 21% of YouTube Shorts were AI-generated, highlighting a significant disparity in content quality between the two platforms.
The prevalence of AI-generated content varies dramatically by category on TikTok. Kids content shows the highest concentration at 57% AI slop overall, with specific hashtags reaching alarming levels—#cartoonkids hit 97%, #cartoons 83%, and #babysong 83%. Science and Education (35%), Health (33%), and History (33%) categories are also heavily affected, likely because they rely on visual illustrations and voiceover narration. In contrast, categories requiring on-camera presence or physical demonstration show significantly lower rates: Fashion (1.3%), Music (1.5%), and Fitness (1.6%).
TikTok has implemented some measures to address the problem, including labeling 1.3 billion videos as AI-generated by November and providing user controls to filter AI content. However, the default experience for new users still heavily favors AI-generated videos. YouTube has similarly rolled out detection systems and monetization policy changes, but the actual impact of these interventions on what users see remains unmeasured.
- Current platform interventions appear insufficient to prevent AI slop from dominating new user feeds
Editorial Opinion
The report exposes a troubling reality: TikTok's recommendation algorithm appears optimized for engagement with AI-generated compilations at the expense of human creativity. For content creators and brands, this represents a deteriorating competitive landscape where cheap, automated content may outperform genuine work. The concentration in children's content is particularly alarming, subjecting young viewers to low-quality AI compilations at scale. Without more aggressive platform interventions beyond labeling and optional filters, TikTok risks becoming a platform where algorithmic recommendations systematically favor automation over artistry.
