Video editing platform Kapwing has published a detailed research report examining the growing presence of low-quality AI-generated videos on YouTube. The study highlights that over 20% of videos recommended to brand-new users fall into what researchers label as “AI slop” — content created primarily to exploit platform algorithms rather than deliver meaningful value to viewers.

Kapwing defines “AI slop” as auto-generated, low-effort videos that are mass-produced to maximize clicks, watch time, and ad revenue. Unlike high-quality AI-assisted videos, these uploads typically rely on recycled visuals, synthetic narration, and minimal editorial oversight. To measure the scale of the issue, researchers created a fresh YouTube account and analyzed the first 500 recommended videos, discovering that roughly 21% matched the AI slop classification.

The report also highlights how this content has become financially significant. One of the most prominent examples cited is India’s “Bandar Apna Dost,” an anthropomorphic monkey channel that has accumulated over two billion views and is estimated to earn more than $4 million annually. The data suggests that YouTube’s recommendation engine often amplifies such content due to strong engagement metrics, regardless of quality.

Geographically, South Korea emerged as the largest consumer of AI slop content with more than 8.4 billion views, followed by Pakistan and the United States. Interestingly, Spanish channels were found to generate the highest subscriber growth, indicating that language and regional dynamics play a role in how algorithmic content spreads.

Looking ahead, Kapwing’s findings raise broader questions about the future of online media ecosystems. As generative AI tools become cheaper and easier to use, platforms may see an increasing flood of automated content. The study reinforces concerns tied to the “Dead Internet Theory,” where synthetic media overwhelms authentic human-created content, reshaping how users discover and trust information online.

While platforms like YouTube continue to refine moderation and ranking systems, the report suggests that economic incentives still favor volume-driven automation. Without structural changes to discovery algorithms, AI slop is likely to remain a persistent feature of the digital video landscape.