Kapwing: 59% of New TikTok Feeds Are AI Slop
Key insights
- Kapwing found 59% of videos recommended to new TikTok accounts qualify as AI slop, three times higher than YouTube's rate.
- Kids' content on TikTok shows the highest AI saturation, with the #CartoonKids tag reaching 97% AI-generated videos.
- TikTok's passive opt-out toggle is insufficient per Kapwing, which calls for active platform-level moderation of AI content.
Why this matters
Platform defaults shape how users first encounter AI-generated content at scale, and a 59% AI content rate at first scroll means new TikTok users are being conditioned to treat synthetic media as normal before they have set any preferences. For founders building AI content tools, TikTok's permissive algorithmic amplification is effectively a distribution channel, but it risks triggering platform crackdowns or regulatory action that could reverse that advantage quickly. YouTube's more aggressive enforcement posture, including channel crackdowns and built-in viewer controls, signals a platform policy divergence that will force content-tool builders to align with whichever moderation regime dominates.
Summary
Kapwing finds 59% of videos shown to new TikTok accounts are AI slop, and the saturation grows as users scroll deeper. The study created fresh TikTok accounts and manually evaluated popular categories, finding TikTok delivers roughly three times more AI content than YouTube to new users.
The numbers are starker in specific niches: kids' content hits 57% AI saturation overall, and within the #CartoonKids tag specifically, 97% of videos were identified as slop.
Essentially: (Kapwing, TikTok) are divided on whether an opt-out toggle is sufficient or active platform enforcement is required.
- TikTok offers users a setting to see less AI content, but Kapwing argues passive controls are not enough.
- YouTube has moved more aggressively, with channel crackdowns and built-in viewer-side time-limit tools.
- AI content density on TikTok increases the deeper a new user scrolls through the For You Page.
A feed nearly 60% synthetic at first contact sets a troubling baseline for how new users learn to read algorithmic recommendations.
Potential risks and opportunities
Risks
- TikTok faces EU Digital Services Act scrutiny if regulators treat a 59% AI-content default as a systemic risk to information integrity for new users.
- Children's content creators lose algorithmic reach as the #CartoonKids tag saturates at 97% AI, crowding human-made kids' video out of recommendations entirely.
- Advertisers running brand-safety audits targeting AI slop could pull spend from TikTok's For You Page inventory, pressuring CPMs if the 59% figure gains sustained mainstream press attention.
Opportunities
- Kapwing gains credibility as an independent AI content auditor and can expect inbound from platforms, regulators, and brands seeking category-level AI saturation breakdowns.
- YouTube can accelerate advertiser wins by publishing its own AI-content rate data, positioning its three-times-lower baseline as a premium brand-safe environment against TikTok.
- AI content detection vendors such as Hive Moderation and Reality Defender have a direct commercial case to pitch TikTok and rival short-form platforms on automated slop filtering at scale.
What we don't know yet
- Whether TikTok's 'see less AI content' toggle measurably reduces AI recommendations or only labels them, and by how much.
- How Kapwing operationally defined 'AI slop' across categories where AI assistance is ambiguous, such as scripted voiceover versus fully synthetic video generation.
- Whether the three-times gap versus YouTube persists for established accounts or is specific to the cold-start For You Page experience.
Originally reported by tubefilter.com
Read the original article →Original headline: Kapwing Study: ~60% of TikTok For-You Videos Shown to New Accounts Are AI-Generated Slop — 3× the Rate on YouTube Shorts