A new study has found that more than 20 per cent of videos recommended by YouTube to new users are “AI-slops”. This term refers to low-quality videos created using artificial intelligence. The research was conducted by video-editing company Kapwing, which reviewed 15,000 of the world’s most popular YouTube channels. These included the top 100 channels from each country.

The study showed that 278 of these popular channels publish only AI slop. These videos usually focus on attracting views and clicks but offer little or no meaningful content.

Massive Views and Big Money

Despite their poor quality, AI slop channels attract huge audiences. Together, they have crossed 63 billion views and gained around 221 million subscribers. According to Kapwing’s estimates, these channels earn nearly $117 million (£90 million) every year through advertising and monetisation.

Algorithm Pushes Low-Quality Content

To test YouTube’s recommendation system, researchers created a new account and tracked the first 500 videos suggested to it. They found that 104 of these videos were AI slop. Nearly one-third of the recommendations fell under a wider category called “brainrot”, which includes AI slop and other low-effort videos made mainly to hold attention and generate revenue.

A Fast-Growing Global Industry

The study highlights the rapid growth of this type of content across major social media platforms such as YouTube, X, and Meta. These videos are often repetitive, context-free, and designed for global audiences. A Guardian report noted that nearly 10 per cent of YouTube’s fastest-growing channels now fall into the AI slop category, even as the platform claims to act against “inauthentic content”.

Strong International Reach

Kapwing found that AI slop channels have followers across the world. In Spain, around 20 million people, almost half the population, follow trending AI-driven channels. Egypt has about 18 million followers, while the United States has 14.5 million and Brazil around 13.5 million.

True Scale Still Unclear

However, experts say it is difficult to measure the full impact of AI slop on YouTube. The platform does not publish detailed data on yearly views or the share generated by AI content. As a result, the exact scale of AI-generated low-quality videos on YouTube remains uncertain.