AI slop makes up 60% of TikTok feeds, study finds

Almost 60% of videos served to new TikTok users are AI-generated slop, according to a new report from San Francisco video editing company Kapwing. The figure is three times the saturation rate found on YouTube.

The findings, reported by Kapwing, reveal how thoroughly algorithmic feeds have been colonised by low-quality synthetic content.

Children's content worst affected

The problem becomes even worse when it comes to the younger generation. According to Kapwing, the "Kids" category is characterised by the highest density of AI slop. The hashtag #cartoonkids is "almost entirely made of slop, with only three of the 100 videos being human-created".

When children browse TikTok looking for educational or entertaining human-made content, they often end up seeing AI-created material on their feeds instead. Once a user shows interest in AI-created videos, the platform begins serving more of them, a pattern researchers describe as a vicious cycle.

Experts warn of developmental risks

Experts have warned that this exposure carries developmental consequences. Low-quality, "brain-melting" AI material could endanger children's cognitive development. At the same time, photorealistic deepfakes are facilitating the spread of misinformation and political propaganda across platforms.

Other platforms also affected

Other social networking sites have not been spared. Facebook and Instagram, both owned by Meta, have become unrecognisable wastelands, according to user comments. These platforms are filled with strange AI-generated images, including depictions of poor children without legs or arms begging on roads, and video footage showing anthropomorphic cats killing their young in meat grinders.

Platforms respond differently

TikTok introduced controls in November allowing users to increase or decrease the amount of AI content shown to them. YouTube, meanwhile, has started marking AI-generated videos but has not yet altered its recommendation algorithm.