YouTube’s public stance on cracking down against low-quality, mass-produced AI videos may sound firm, but a new study suggests the platform’s recommendation system is still pushing large amounts of such content to users, including first-time viewers. The findings raise fresh questions about how effective YouTube’s policies really are, especially at a time when AI-generated videos are becoming easier and cheaper to produce at scale. According to a report by video-editing company Kapwing, more than one in five videos recommended to new YouTube accounts can be classified as what it calls “AI slop”. The researchers analysed content from 15,000 of the world’s most popular YouTube channels to understand how widespread this trend has become and how much engagement these videos are pulling in.
AI slop YouTube channel from India made $4.25M with Bandar Apna Dost videos
The scale of the issue becomes clearer when looking at the numbers. Kapwing found that 278 of the analysed channels were exclusively uploading AI slop videos. Together, these channels have clocked a staggering 63 billion views and amassed around 221 million subscribers. Despite YouTube’s rules stating that such low-quality AI-generated content is not eligible for monetisation, the report estimates that these channels could collectively be earning about $117 million every year through indirect means.
India features prominently in the findings. The most-viewed AI slop channel identified in the study is ‘Bandar Apna Dost’, which Kapwing says is based in India. The channel alone has crossed 2.4 billion views. Its videos typically feature an AI-generated rhesus monkey with human-like traits, often paired with a Hulk-like muscular character, battling demons in dramatic, repetitive storylines designed to keep viewers hooked. Kapwing estimates that this single channel could be generating around $4.25 million in annual revenue, even if it is not officially monetised under YouTube’s standard ad policies.
To understand how this content reaches viewers, the researchers also tested YouTube’s recommendation engine directly. After setting up a fresh account, they tracked the first 500 videos suggested on the home feed. Out of these, 104 were labelled as AI slop. Around one-third of the remaining recommendations fell into a broader category the researchers described as “brain rot”, referring to low-effort, highly repetitive content optimised mainly for clicks and watch time.
The report paints a picture of a fast-growing, loosely organised ecosystem built around generative AI tools. At one level are creators who churn out large volumes of AI-generated videos using freely available software. At another are individuals, often operating in grey areas, who sell courses and “guaranteed viral” strategies teaching others how to replicate this model and farm engagement.
While YouTube has always hosted a mix of high-quality and questionable content, AI slop stands out because of how quickly it can be produced and scaled. A single creator, or even a small team, can upload dozens of similar videos daily with minimal cost, something that was not feasible in the pre-AI era.
YouTube has already started blocking fake AI content
The growing frustration is not limited to YouTube. Across platforms like Instagram and X, users have been complaining that their feeds are increasingly cluttered with strange, repetitive, and low-quality AI-generated posts. In response, platforms have tightened rules and relied more on takedowns. Earlier this month, YouTube reportedly blocked two large channels that were uploading fake AI-generated movie trailers, a move seen as part of this broader cleanup effort.
At the same time, there is an apparent contradiction in how big tech companies view AI content. During an earnings call in October, Meta CEO Mark Zuckerberg spoke openly about AI helping create “yet another huge corpus of content” for Facebook and Instagram, arguing that easier creation and remixing would boost sharing and engagement. YouTube, too, has leaned into this future by integrating Veo 3, Google’s latest AI video generator, directly into Shorts, allowing users to create AI-powered vertical videos within the app.
Responding to Kapwing’s findings, YouTube defended its approach. “Generative AI is a tool, and like any tool it can be used to make both high- and low-quality content,” a company spokesperson said, as quoted by The Guardian. The platform added that it remains focused on connecting users with high-quality content and that all videos must comply with its community guidelines, regardless of how they are made.
– Ends
