New Study Reveals 25% of YouTube Videos for New Users Are Low-Quality AI Content

New research from Kapwing reveals that more than one in five videos recommended to new YouTube accounts are low-quality, AI-generated content, often referred to as “AI slop.” This alarming trend highlights a significant shift in the platform’s content ecosystem, raising concerns about the quality of material being fed to new users.

The study analyzed the performance of various YouTube channels and the experience of users with newly created accounts. Researchers found that of the first 500 videos suggested to a new account, 104 were classified as AI slop, accounting for over 20% of the recommendations. This finding indicates that YouTube’s recommendation system is not merely an accident but rather a systematic approach that exposes new users to low-quality content regularly.

The Impact of AI Slop on YouTube’s Ecosystem

AI slop is not the only concern. The analysis also revealed that approximately one-third of the first 500 videos viewed by new accounts consist of compulsive, low-quality content described as “brainrot,” regardless of whether it is AI-generated. The implications extend beyond individual user experiences; Kapwing’s review of the top 100 trending YouTube channels in various countries found that numerous high-ranking channels are devoted primarily to AI-generated videos. These channels collectively boast billions of views and millions of subscribers, generating an estimated $117 million in annual revenue.

Content produced by these channels often features surreal or absurd animated shorts lacking clear narratives, as well as bright, looping animations designed for children. Countries such as Spain, South Korea, Egypt, Brazil, and India have emerged as significant markets for these AI-generated videos. Notably, the India-based channel Bandar Apna Dost was identified as the most-viewed channel in the study, with over 2 billion views.

YouTube’s Stance on Generative AI Content

Despite these findings, YouTube has stated that it will not ban generative AI content as long as channels disclose its use. This stance raises critical questions: Why does the platform maintain a neutral position while its recommendation systems are amplifying low-value content? Furthermore, what are the potential consequences if this trend continues, particularly concerning the retention of top creators on the platform?

As AI slop proliferates, YouTube faces a crucial challenge in balancing the interests of content creators and the quality of its recommendations. Users frustrated with the prevalence of AI slop and brainrot may find it beneficial to learn how to refine their recommendations to enhance their viewing experience.

The study emphasizes the urgent need for YouTube to reassess its content guidelines and recommendation algorithms to ensure a healthier ecosystem for all users.