TikTok Floods New Users With AI Slop at Triple the Rate of YouTube, Study Finds

A new study from video editing platform Kapwing reveals that 59% of videos served to a fresh TikTok account's For You feed are AI-generated slop, roughly three times the rate found on YouTube Shorts. The report, published June 21, 2026, manually reviewed 10,742 TikTok videos across 20 categories and ran separate fresh-account tests on both platforms.

In the fresh-account test, 294 of the first 500 TikTok For You videos were AI slop, compared to 104 of 500 YouTube Shorts, or 21%. TikTok had already labeled 1.3 billion videos as AI-generated as of November 2025, according to the report.

Kids Content Dominates AI Slop, Science and Education Also Hit Hard

The Kids category had the highest rate of AI slop at 57% of 2,000 reviewed videos. The tag #cartoonkids topped 97% AI generation among 100 featured videos. Tags #cartoons and #babysong both reached 83%, and #forkids hit 79%.

• Science and Education: 35% AI slop
• Health: 33%
• History: 33%
• Fashion: 1.3% (lowest)
• Music: 1.5%
• Fitness: 1.6%

Categories relying on on-camera presence or physical demonstration had the lowest rates. The study suggests that visual illustration and voiceover narration make Science, Health, and History categories more susceptible to AI-generated content.

Platform Incentives and the Race to Moderate

The disparity between TikTok and YouTube raises questions about each platform's content moderation and algorithmic incentives. TikTok's For You feed prioritizes engagement, which AI-generated content often exploits with eye-catching but low-quality visuals. YouTube has publicly acknowledged its AI slop problem, with CEO Neal Mohan stating in 2025 that the platform would curb low-quality AI content.

Kapwing's findings come as social media platforms grapple with the flood of generative AI content. TikTok's labeling of 1.3 billion videos shows some effort, but the high proportion of AI slop on new accounts suggests the algorithm still surfaces it aggressively. For advertisers and content creators, the data points to a growing trust gap: users may struggle to distinguish real from synthetic, especially in categories like Kids and Education where authenticity matters most.

What comes next is uncertain. Platforms may tighten AI labeling requirements or adjust recommendation algorithms to deprioritize synthetic content. Regulators in the EU and elsewhere are already eyeing AI-generated content rules. For now, the study serves as a baseline: nearly six in ten videos on a new TikTok feed are not made by humans.