A recent analysis by video editing platform Kapwing found that 59% of the videos served to brand-new TikTok users consist of AI-generated content, often characterized by low quality and factual inaccuracies. This phenomenon, dubbed AI slop, is especially prevalent in categories aimed at children and educational topics.
- 59% of videos shown to new TikTok accounts are AI-generated low-quality content
- Children’s and educational content categories are most heavily impacted
- TikTok’s recent tools to limit AI content have not significantly reduced AI slop
What happened
Kapwing conducted an extensive study analyzing 10,742 TikTok videos from 20 different categories as well as the first 500 videos presented to newly created TikTok accounts. The study found that 59% of the videos served to new users featured obvious AI-generated visuals, poorly scripted content, and synthetic voiceovers, a phenomenon Kapwing refers to as AI slop. This rate is approximately three times higher than the same study’s findings for YouTube. Among the worst affected categories is TikTok’s Kids section, where 57% of videos are AI-based, followed by education, science, health, and history content.
The study also uncovered that while some categories such as fitness, music, and fashion have barely any AI-generated videos, tags associated with children’s content—like #CartoonKids and #cartoons—are almost entirely dominated by AI-created clips. These videos often display familiar cartoon characters in nonsensical situations or provide inaccurate educational lessons, with content generated so rapidly that it outpaces what human creators could produce. This saturation of AI content marks TikTok’s initial impression to new users as mainly synthetic and low quality.
Why it matters
The predominance of AI-generated content among new TikTok users poses serious challenges for content accuracy, platform trust, and user safety. In particular, the high volume of AI slop in children’s and educational content is concerning because young viewers may be exposed to misleading information without the critical context to question its validity. Experts have labeled this a form of large-scale misinformation targeted at toddlers, raising ethical and regulatory concerns.
Beyond misinformation, TikTok’s dependency on AI slop for fresh user feeds undermines its reputation as a platform driven by engaging and authentic creator content. This issue exists during the critical period when TikTok’s recommendation algorithm has not yet personalized the experience based on user interaction signals. Although TikTok has introduced user controls to adjust AI content levels and launched AI literacy programs, these measures have not effectively reduced the volume of AI-generated videos presented to newcomers, leaving the problem largely unaddressed.
What to watch next
Monitoring TikTok’s response to the AI slop challenge will be crucial, especially as regulatory scrutiny intensifies. Recently, Florida filed a lawsuit alleging TikTok allowed minors onto the app and misled parents about the content’s nature. Findings from the Kapwing study add another layer to this scrutiny by highlighting how AI-generated junk floods children’s feeds, potentially worsening legal and reputational risks for the platform.
Moving forward, stakeholders will be watching to see if TikTok can improve its AI content moderation and curation, especially in its onboarding experience and kid-focused categories. Enhancements in detection and filtering, stronger regulatory compliance, and a greater commitment to human-generated authoritative content could influence the platform’s ability to maintain user trust and avoid further criticism.