If you have ever felt like your TikTok feed is mostly fake content, you are not imagining it. A new report from Kapwing found that 59% of videos served to a brand-new TikTok account are AI slop. That is roughly three times the rate Kapwing found when it ran the same test on YouTube.
The study, conducted by the video editing platform Kapwing, aimed to quantify the prevalence of AI-generated content on two of the world's most popular short-form video platforms. By creating fresh accounts with no prior viewing history, the researchers sought to understand what new users encounter by default. The results paint a stark picture of the state of content moderation and algorithmic curation in the age of generative AI.
Kapwing built a brand-new account on both TikTok and YouTube, and manually checked the first 500 videos served to each one. On TikTok, 294 of those videos were AI-generated, representing a staggering 58.8% of the feed. On YouTube, only 104 of the first 500 Shorts qualified as slop, putting that platform’s rate at 21%. This threefold difference suggests that TikTok's algorithm is either more permissive of AI-generated content or actively prioritizes it for new users.
The scale of the problem is staggering when you consider that TikTok had already labeled 1.3 billion videos as AI-generated by November of last year. This number reflects only those videos that were voluntarily labeled by creators or detected by TikTok's automated systems. The actual number of AI-generated videos on the platform is likely much higher, as many go undetected or unlabeled. Kapwing also manually reviewed over 10,000 TikTok videos across 20 different content categories to get a fuller picture of where slop tends to cluster.
Which TikTok categories are flooded with AI slop
The study broke down AI content prevalence by category, revealing significant disparities. Kids’ content topped every category, with 57% of the 2,000 videos turning out to be AI-generated. The worst single tag was cartoonkids, where 97 out of 100 featured videos were artificial. This is particularly concerning because children are a vulnerable audience who may not be able to distinguish between real and AI-generated content. The proliferation of AI in kids' content could have implications for child development, media literacy, and safety regulations.
Science and Education, Health, and History followed close behind, each landing between 33% and 35% AI slop. These are categories where animation and voiceover narration tend to replace real demonstration. For instance, many educational channels now use AI-generated images and text-to-speech narration to produce videos quickly and cheaply. While this can increase content output, it also risks spreading misinformation if the AI generates inaccurate or misleading information. In the health category, AI-generated advice could be particularly dangerous if it contradicts established medical guidelines.
On the other end, Fashion, Music, and Fitness were nearly untouched, each sitting below 2%, likely because those formats rely heavily on real, on-camera presence. These categories require genuine human performance, body language, and physical demonstration, which AI currently cannot replicate convincingly. This suggests that AI slop thrives in categories where visual and audio elements can be easily synthesized, such as static images, simple animations, and generic voiceovers.
Even though TikTok has rolled out tools for users to dial back AI content in their feeds, this study suggests that what shows up by default still leans heavily towards AI. These tools include options to mark content as “not interested” and a new feature that allows users to filter out AI-generated posts. However, these are opt-in mechanisms, meaning the burden of filtering slop from substance largely falls on the viewer. Many users may not even be aware that such tools exist or may not bother to use them. The study underscores the need for platform-level interventions to ensure that users are not overwhelmed by low-quality, synthetic content from the moment they join.
The findings also raise questions about the effectiveness of current labeling policies. TikTok requires creators to label AI-generated content that is photorealistic or depicts real people in fake scenarios. However, many AI-generated videos fall into gray areas—such as cartoon animations or stylized imagery—that may not trigger the labeling requirement. As AI generation tools become more sophisticated, distinguishing between human and machine-made content will only become harder.
YouTube, on the other hand, has taken a more cautious approach. Its Content ID system and community guidelines place stricter limits on automated content, and algorithmic recommendations for Shorts appear to favor original human-created material. This may explain the lower rate of AI slop on the platform. However, as competition for short-form video attention intensifies, these standards may come under pressure.
The broader implications of this study are significant. AI-generated content is already reshaping the information ecosystem, and platforms must decide how to handle the flood of synthetic media. The rise of AI slop threatens to degrade user experience, spread misinformation, and undermine trust in online content. Without robust detection and moderation, users—especially new ones—will be fed a diet of machine-made videos that look real but lack substance. The study from Kapwing serves as a wake-up call for both platforms and users to take AI content awareness more seriously.
As generative AI continues to advance, the challenge will only grow. Multimodal models can now create realistic video, audio, and text from a single prompt. The line between authentic and AI-generated content is blurring. For platforms like TikTok, which thrive on viral, low-effort content, AI slop is a lucrative but dangerous trend. For users, the responsibility to filter out noise and find genuine content grows heavier each day.
Source: Digital Trends News