TikTok AI – A new Kapwing study found that 59% of videos served to a fresh TikTok account are AI-generated “slop,” about three times the rate seen on YouTube Shorts. The testing and a larger category review point to kids’ content as the biggest hotspot.
For anyone who has opened TikTok and felt that familiar frustration—scrolling past clips that look oddly polished, generic, and synthetic—this latest report lands like proof.
Kapwing says its test found that 59% of videos served to a brand-new TikTok account are AI slop. In the same style of check on YouTube, the rate came out far lower.
To measure it, Kapwing created a fresh account on both platforms and manually reviewed the first 500 videos shown. On TikTok, 294 of those videos were AI-generated. On YouTube, only 104 of the first 500 Shorts qualified as slop, for a rate of 21%.
The scale is harder to ignore when you zoom out. By November, TikTok had already labeled 1.3 billion videos as AI-generated. Kapwing also went beyond the initial feed test, manually reviewing more than 10,000 TikTok videos across 20 content categories to see where the problem clusters.
The results were brutally clear: kids’ content was the most flooded. Kapwing reports that 57% of the 2,000 videos it reviewed in kids’ categories were AI-generated. The worst single tag was #cartoonkids, where 97 out of 100 featured videos were artificial.
Science and Education, Health, and History followed close behind, each landing between 33% and 35% AI slop. Kapwing points to a practical reason these categories may attract synthetic content: animation and voiceover narration can stand in for real demonstrations.
On the other end of the spectrum, categories such as Fashion, Music, and Fitness were nearly untouched, each sitting below 2%. The report frames that gap as likely connected to formats that rely heavily on real, on-camera presence.
TikTok has also rolled out tools meant to let users dial back AI content in their feeds. Even with those controls available. the study’s core takeaway is that the default reality of what users see still leans heavily toward AI. For now, the burden of separating slop from substance largely falls on the viewer’s own filtering and patience.
TikTok YouTube Shorts Kapwing AI slop generative AI content moderation misinformation creator economy social media trends #cartoonkids
So 59% is AI garbage? Cool cool, guess that’s why my feed keeps looking like the same dumb ad with different words.
I don’t even know if that “study” is real. Like they made a fresh account and checked 500 videos… I could get totally different stuff than them on TikTok. Also YouTube Shorts is way worse for me so idk.
Cartoon kids at 97%?? That sounds like something somebody should be arrested over, not just a report. If AI voices are tricking kids then that’s like… crime-adjacent. And TikTok says it’s labeled but I swear half the time it’s still hard to tell.
“Controls” to dial back AI content… sure, I tried those settings once and it didn’t change anything. TikTok keeps feeding me the polished “too perfect” videos and then acts like it’s my fault for not watching the real ones. Also how does the Fashion/music stuff only get below 2%—because those people actually film stuff? or because the study just didn’t look long enough.