Voicing a pertinent concern for the present day, Neal Mohan, CEO of YouTube, acknowledged in his annual letter published last month, that “the lines between technology and creativity are blurring”. In the same letter he problematised ‘AI slop’, saying it is a priority for YouTube in 2026 to monitor and regulate slop content — a new word (recognised by Merriam-Webster as the word of the year for 2025), and a new concept that the digital world is currently witnessing.
Mohan also mentions that this year YouTube will also be introducing new AI features to allow creators to use their likeness in the content they create. An AI-coded conundrum and contradiction all rolled into one — not only YouTube, but other social media sites are also facing an outpouring of AI slop, or low-quality AI generated content, and are garnering views and followers as well.
A Kapwing research released in November of last year found that globally, more than 20% of the content that YouTube algorithm shows us can be classified as AI slop. These AI slop channels collectively make about $117 million each year, with 63 billion views and 221 million subscribers at the time of this survey. The Kapwing study also involved the researchers creating a dummy account to monitor outcomes of interactions with the YouTube algorithm.
It was noted that of the 500 videos suggested to the user, 104 were AI slop videos, and one-third of the 500 were brainrot content (also categorised under AI slop). Another survey done by The Guardian in 2025 found that nearly 10% of YouTube’s fastest growing channels were AI slop, racking up millions in views.
Economics of Absurdity
The study found that the AI slop channel collecting the most views, called Bandar Apna Dost, with 2.65 billion views on YouTube, and over 3 million subscribers. The channel features videos with AI-generated characters based on a human-like monkey and the Incredible Hulk, in various absurd settings like aeroplanes, funerals, and hospitals, among others.
Other channels worth mentioning are the Pouty Frenchie based in Singapore with over 6.65 million subscribers which follows the activities of a French bulldog, and Pakistan-based The AI World with over 800K subscribers, where one can find video
recreations of village life during devastating floods, as well as village life in the desert regions.
More recently, with the mad frenzy around the release of the last season of popular Netflix show Stranger Things in December 2025, a large number of AI slop channels took to making AI generated videos of the show’s characters. Ridiculous videos with the red, veiny, rather terrifying monster of the series, Vecna dancing in the market donning a lungi, fan favourite characters drinking chai at a roadside stop in Indian garb, and other innovative and equally unrealistic scenarios.
From Shrimp Jesus to Vecna
A massive suspension of disbelief and a tolerance for the farcical is imperative to be able to sit through a ‘slop’ reel or video. What has spun the problem further out of control is the fact that images and videos can be generated with artificial intelligence with a simple text prompt alone.
The first real example of AI slop became viral in 2024 with ‘Shrimp Jesus’. A visual that became widely shared for no apparent rhyme or reason was an AI-warped image of Jesus with the bottom half of a shrimp. Various iterations of this image were subsequently created and circulated by others, until the trend made way for more advanced or skilled creators creating this significant niche on the internet.
Another common favourite in the AI slop world are the cat videos, predicting human-like cats in decidedly human settings partaking in human activities, like cooking in the kitchen, working out, hunting, getting arrested, having affairs and more. A popular channel catering to this content is the Super Cat League on YouTube, which boasts nearly 4 billion subscribers. Facebook is also crowded with AI generated videos of cats in adult clothes walking on their hind legs.
It is also important to point out that these videos are not restricted to comedy content alone, but often even portray morbid and violent scenarios that might have been flagged if the content was filmed and not AI generated.
The ease of using AI to create images and videos has understandably exacerbated this problem. Not only is this content often pointless and bordering on the impossible, it also throws a significant wrench in genuine creatives trying to find their visibility and space on the internet space.
According to a McAfee consumer survey conducted on American users in 2025, Americans spend almost 114 hours on average in a year, trying to determine whether a piece of content is authentic or AI generated.
Similar studies do not exist for other populations, but the influence of AI, and the reach is evident from the aforementioned channels and where they are based. The question now is, will YouTube’s upcoming measures serve to regulate the sheer scale and quantity of AI slop being generated on a daily basis? Or will the advent of AI, its use, and the multiple free range training sources for AI win over, bringing us even more unoriginal art and slop this year.