TikTok Serves 60% AI-Generated Content to New Users, Report Finds

Synthetic Content Dominates TikTok's Algorithm

Nearly 60 percent of videos appearing on TikTok's algorithmic For You page for new users are now AI-generated, according to research published by San Francisco video editing company Kapwing. The figure represents a threefold increase compared to YouTube, where new users encounter roughly 20 percent synthetic content.

The research, first reported by Futurism, examined content distribution patterns across major social platforms and found TikTok's algorithm particularly aggressive in serving AI-generated material. Once an account demonstrates interest in synthetic content, the platform's recommendation system amplifies the pattern, delivering even higher concentrations of AI-generated videos.

Children's Content Shows Highest Concentration

Kapwing's analysis identified children's categories as the most heavily affected segment. The hashtag #cartoonkids consisted almost entirely of synthetic material, with only three human-created videos among 100 examined. This concentration raises questions about the exposure of young users to algorithmically generated content during critical developmental periods.

The platform is not alone in confronting this challenge. Meta's Facebook and Instagram have experienced similar shifts, with feeds increasingly populated by synthetic images and videos that blur the line between human creativity and machine output.

Why it matters

The dominance of AI-generated content on platforms used by hundreds of millions represents more than a content quality issue. Experts warn that sustained exposure to synthetic material may affect cognitive development in young users, while photorealistic deepfakes create new vectors for misinformation and political manipulation. As detection becomes harder—even leading deepfake researchers report difficulty distinguishing real from synthetic—platforms face mounting pressure to implement effective controls without clear technical solutions.

Platform Responses Remain Limited

TikTok announced in November that users would gain controls to adjust the volume of AI-generated content in their feeds. Jade Nester, the platform's European director of public policy for safety and privacy, told The Guardian the feature acknowledges that some users appreciate synthetic content while others prefer to limit exposure.

YouTube has introduced labeling requirements for AI-generated material but stopped short of modifying recommendation algorithms or monetization eligibility based on content origin. The approach reflects the technical difficulty platforms face: as synthetic content quality improves, distinguishing between human and machine creation becomes increasingly complex.

Hany Farid, described as the world's leading deepfake expert, told the New York Times he has stopped trusting his own visual assessment—a stark indicator of how rapidly synthetic media capabilities have advanced beyond detection methods.

These findings were first reported by Futurism, based on research conducted by Kapwing.