YouTube’s crackdown on mass-produced, low-quality AI-generated content appears to be falling short, with a new study claiming that the platform’s own recommendation algorithm continues to surface ‘AI slop’ to new users.
Over 20 per cent of the videos suggested to new users by YouTube’s algorithm is AI slop, according to a report published by Kapwing, a video-editing company, which analysed 15,000 of the world’s most popular YouTube channels to find out which ones are churning out AI slop content and how much in views and revenue those videos are generating.
The researchers found that 278 out of 15,000 YouTube channels had uploaded only AI slop videos, drawing a total of 63 billion views and recording 221 million subscribers. The most-viewed AI slop channel on YouTube (over 2.4 billion views) is called ‘Bandar Apna Dost’ and is based in India, as per Kapwing. AI-generated videos uploaded by this YouTube channel feature an anthropomorphic rhesus monkey and another Hulk-like muscular character fighting demons.
While AI slop videos are not eligible for monetisation under YouTube’s existing policies, Kapwing has estimated that such YouTube channels could collectively rake in about $117 million revenue each year with Bandar Apna Dost alone estimated to generate about $4.25 million yearly revenue.
After creating a new account on YouTube, the researchers found that 104 of the first 500 videos recommended to its feed were AI slop and another one-third of them could be categorised as ‘brain rot’.
Kapwing’s findings offer a rare glimpse into a rapidly growing, semi-structured industry that is using generative AI tools to farm engagement. It also includes another tier of people, mostly scammers, offering paid tips and courses on how to make viral, AI-generated content.
There is a vast sea of authentic and inauthentic content on YouTube, but AI slop is particularly concerning because it can be mass-produced at scale using an array of freely available AI tools.
From Instagram to X and YouTube, social media users have been increasingly complaining that their feeds are now saturated with AI slop. In response, platforms have moved to curb the spread of such low-quality content by tightening their policies and relying on takedown mechanisms. Earlier this month, YouTube reportedly blocked two massive channels that peddled fake, AI-generated movie trailers.
However, big tech companies also view AI-generated content as the future of social media. During an earnings call in October, Meta CEO Mark Zuckerberg said that the Facebook and Instagram parent will “add yet another huge corpus of content” to its recommendations system as AI “makes it easier to create and remix” work that gets shared online.
YouTube has also integrated Veo 3, Google’s latest AI video generator, directly into Shorts, allowing users to create AI-generated vertical videos within the platform.
Commenting on Kapwing’s report, YouTube said, “Generative AI is a tool, and like any tool it can be used to make both high- and low-quality content.” “We remain focused on connecting our users with high-quality content, regardless of how it was made. All content uploaded to YouTube must comply with our community guidelines, and if we find that content violates a policy, we remove it,” a company spokesperson was quoted as saying by The Guardian.
Amidst the proliferation of AI-generated content on the internet in the past 12 months, American dictionary-maker Merriam-Webster announced that its word of the year for 2025 is ‘slop’. It has defined the term as “digital content of low quality that is produced usually in quantity by means of artificial intelligence.”
“The flood of slop in 2025 included absurd videos, off-kilter advertising images, cheesy propaganda, fake news that looks pretty real, junky AI-written books, “workslop” reports that waste coworkers’ time… and lots of talking cats. People found it annoying, and people ate it up,” it said.
“It’s such an illustrative word. It’s part of a transformative technology, AI, and it’s something that people have found fascinating, annoying, and a little bit ridiculous,” Merriam-Webster’s president, Greg Barlow, was quoted as saying by The Associated Press.
Other tech-related words have also been picked as word of the year by dictionaries. For instance, Oxford Dictionary chose ‘ragebait’ – “online content deliberately designed to elicit anger or outrage by being frustrating, provocative, or offensive, typically posted in order to increase traffic to or engagement with a particular web page or social media content”.
Collins Dictionary, on the other hand, went with ‘vibe coding’ – a term coined by famed AI researcher Andrej Karpathy, it refers to “the use of artificial intelligence prompted by natural language to write computer code.”