One in Three TikToks Shown to New Users Is Low-Quality AI Slop, Study Finds

When a new user opens TikTok for the first time, their “For You” feed is already flooded with low-effort, AI-generated content. According to a detailed analysis by Kapwing, 59% of the first videos served to a fresh TikTok account qualify as “AI slop” — careless, low-quality videos produced with minimal human oversight.

That figure is nearly three times higher than on YouTube Shorts, where the same methodology found AI-generated content in about 21% of videos shown to new users.

---

What Counts as “AI Slop”?

Kapwing defines AI slop as content that is “careless, low-quality” and created primarily through automated tools.

These videos often feature:

• Poor or glitchy animation;
• Misspelled text and misshapen lettering;
• Factual inaccuracies and historical distortions;
• Sensationalized or misleading claims.

The problem is especially acute in categories aimed at children. 57.4% of videos in TikTok’s “Kids” category were identified as AI slop. That’s more than half of the content being pushed to young audiences — material the researchers warn could deliver “toddler AI misinformation at an industrial scale” and negatively affect developing brains during critical periods of growth.

---

Widespread Saturation Across the Platform

TikTok as a whole has become heavily saturated with AI-generated videos. By November 2025, the platform had already labeled 1.3 billion videos as containing AI-generated content.

The highest concentrations of low-quality AI slop appear in:

• Kids — 57.4%;
• Science & Education — 35.0%;
• Health — 33.8%;
• History — 33.5%.

In contrast, categories like Fitness, Music, and Fashion remain overwhelmingly human-made.

---

User Backlash and the New Normal

Many users are actively pushing back. Negative comments calling out AI-generated content often receive more likes than the original videos themselves.

People have highlighted bizarre mistakes, such as AI misidentifying dancers or producing nonsensical summaries.

Despite the complaints, the volume of AI slop continues to grow, and it appears to be settling into a new status quo on the platform.

---

The “Third Phase” of Social Media

Meta CEO Mark Zuckerberg has described the current moment as the third phase of social media. The first phase was dominated by personal content from friends and family. The second phase shifted toward creator and influencer content. Now, he argues, AI-generated and AI-summarized material is becoming a major new category that platforms will actively promote and fill feeds with.

TikTok’s experience suggests this shift is already well underway — and happening faster, and with lower quality thresholds, than many anticipated.

Also read:

• First Impressions After 24 Hours with Fable: The Token-Rich World and What It Means for Everyone
• Google Phases Out AMP Cache: A Long-Overdue but Welcome Cleanup
• The Duolingo Owl Just Got Even Scarier: Now It Can Block TikTok, Twitter, and Your Entire Distraction Empire
• How Brands Can Break Into AI Search Results: Everything You Wanted to Know About GEO

---

What This Means

The Kapwing report highlights real risks:

• Children are being exposed to large volumes of low-quality, sometimes factually incorrect content.
• Educational and health-related topics — areas where accuracy matters most — are among the most affected.
• The signal-to-noise ratio on short-form video platforms is declining, making it harder for high-quality creators to break through.

While TikTok has introduced labeling tools, user controls for AI content volume, and even a fund for AI literacy, the sheer scale of AI slop suggests these measures are struggling to keep pace with the flood of generated content.

As AI video tools become even more accessible and powerful, the question is no longer whether platforms will be filled with synthetic media — but how much low-quality “slop” users (and especially children) will have to wade through to find anything worthwhile.