Show summary Hide summary
YouTube feeds are filling up with cheaply made, AI-generated clips, and a new analysis finds the phenomenon is far larger than casual viewers may notice. The study, conducted by video-editing platform Kapwing, maps the reach of what it calls “AI slop” and reveals striking numbers about views, subscribers and how often these clips appear in recommendation feeds.
Scope of the Kapwing study and what the researchers measured
Kapwing examined the top 100 YouTube channels in each country. That approach produced a broad dataset that highlights where AI-generated content is concentrated.
Anglo-Saxon burial reveals “unprecedented” secrets: experts stunned by 1,400-year-old grave mysteries
What Your Instinctive Tree Choice Reveals About Your Personality—Experts Explain
- Sample size: 15,000 channels reviewed.
- Identified AI-heavy channels: 278 channels classified as predominantly AI content.
- Combined impact: These channels account for over 63 billion views and roughly 221 million subscribers.
- Estimated annual revenue: Around $117 million across those channels.
How Kapwing tested YouTube’s recommendations
The company also created a brand-new YouTube account to observe real-time recommendations. The experiment focused on the first 500 videos suggested to that fresh profile.
- AI presence in recommendations: 104 of the first 500 suggested videos were AI-generated (about 21%).
- Shorts in the feed: A genre called “brainrot” made up 33% of the first 500 Shorts shown.
Countries and channels where AI videos dominate
The distribution of AI-generated clips is uneven across the globe. A handful of countries and channels are driving a disproportionate share of views.
- Spain: Leading in subscriber totals for AI-heavy channels, with roughly 20.22 million combined subscribers.
- South Korea: AI-heavy channels here have amassed about 8.45 billion views.
- United States: Several prominent channels flagged by the study are U.S.-based or appear in U.S. rankings.
Top-performing AI channels and their reach
- Most-viewed AI channel: Bandar Apna Dost (India) — approximately 2.07 billion views and estimated annual earnings of about $4.25 million.
- Notable subscriber totals: Channels cited in the U.S. list include accounts with several million followers, further amplifying reach.
Why recommendation systems amplify AI-made clips
Recommendation algorithms favor content that drives quick engagement. That creates fertile ground for short, eye-catching AI videos.
- Low-cost production lets creators pump out high volumes of shorts.
- Feeds prioritize immediate watch time and repeat views over creative origin.
- Children’s accounts and family devices often surface these clips more often.
Result: Algorithmic signals can push AI-generated clips into wide circulation fast.
Public response and viral examples raising alarms
Reaction on social platforms has been blunt. Many users express frustration that recommendation systems expose children to a flood of synthetic, low-quality clips.
One widely shared post highlighted a recent short that mixes a childlike aesthetic with surreal and destructive imagery. The clip became one of the top-viewed videos globally for the week, surpassing 100 million views in six days.
Among commenters, common themes include:
- Concern about children’s feeds being dominated by AI-made shorts.
- Criticism that platforms reward volume over authentic creative work.
- Calls for platforms to improve detection and labeling of synthetic content.
What creators and parents are saying
Creators who invest time in original work report feeling sidelined. Parents and caretakers warn that a casual glance at a child’s Shorts feed often reveals a steady stream of synthetic clips.
Social commentators argue that the core challenge is not only detection, but how systems rank and monetize low-effort videos.
