Study Finds 60% of TikTok Videos for New Users Are AI-Generated

A recent analytical report by Kapwing has revealed that new users on TikTok are frequently exposed to low-quality content or videos entirely generated by artificial intelligence, a phenomenon referred to as "AI Slop." This trend raises growing concerns regarding the quality of material that users encounter from the very start of their experience on the platform.

The study examined 10,742 videos across 20 popular TikTok categories and analyzed the first 500 videos appearing on the "For You" page of a completely new account. Findings showed that 294 out of these 500 videos, approximately 59%, were classified as AI-generated content or low-quality compilations clearly relying on AI tools.

According to the report, the proportion of AI-generated content presented to new TikTok users is roughly three times higher than that on YouTube, where the figure was about 21% in the same study. Researchers interpret this disparity as indicating that the initial impression for new TikTok users largely depends on automated content before the recommendation algorithms begin to tailor material to individual interests.

Children’s Content Most Affected by AI Generation

The study identified the children’s category as having the highest saturation of AI-generated content, with 57% of the analyzed videos in this segment falling into that classification, surpassing all other categories. The science and education category ranked second at 35%, followed by health at approximately 34%, and historical content at a similar rate. These categories typically require high accuracy, intensifying concerns about the spread of inaccurate or misleading information.

Hashtag analysis revealed that the #CartoonKids tag was among the clearest examples of this trend, with 97 out of 100 videos examined being AI-generated, leaving only three videos appearing to be human-produced. Other children-focused hashtags also showed high AI content percentages, including #cartoons and #babysong at 83% each, and #forkids at 79%.

Educational Risks and Misinformation Concerns

The report highlighted that many videos feature familiar cartoon characters acting in illogical scenes, educational lessons containing factual errors, synthetic voices, and inconsistent animations. Dr. Dana Suskind, a pediatric professor at the University of Chicago, warned that this represents "widespread misinformation for young children through AI," emphasizing that children lack the capacity to distinguish accurate from false information.

The issue extends beyond children’s content, as the study also found that education, science, health, and history categories are among the most saturated with AI-generated material, potentially facilitating the dissemination of inaccurate information among users.

Role of TikTok’s Algorithms in Content Exposure

The report focuses on the period before content personalization occurs, when TikTok’s algorithm has no behavioral data on a new account. In this phase, the platform relies on initial predictions to select videos for the user. The study suggests this results in AI-generated content serving as the platform’s "default first impression" for new users.

Researchers consider this a significant challenge for TikTok, a platform historically known for its ability to provide precise recommendations that quickly adapt to user preferences.

TikTok Faces Increasing Pressure and Legal Scrutiny

In November 2025, TikTok introduced tools allowing users to increase or decrease the amount of AI-generated content they see and invested in AI awareness programs. However, the report indicates these measures have not significantly reduced the exposure of low-quality content to new accounts.

The company is also under growing legal scrutiny regarding child protection. The state of Florida filed a lawsuit against TikTok under state laws concerning children's use of social media, accusing the platform of permitting minors to use the app and presenting a misleading portrayal of the nature of content available to them.