Low-quality artificial intelligence content has taken over major social media platforms at an alarming rate, with new data showing more than 20% of videos shown to new YouTube users now classified as “AI slop” by researchers.
The troubling findings emerged from a comprehensive study by Kapwing, a video editing firm, which examined 15,000 of YouTube’s most successful channels. The analysis covered the top 100 channels in every nation worldwide, revealing that 278 channels exclusively publish cheaply made, algorithm-gaming videos created solely to generate views and advertising dollars.
These operations have achieved massive scale despite their questionable quality. The identified channels collectively command over 63 billion views and maintain 221 million subscribers. Industry analysts estimate these content farms pull in approximately $117 million annually, equivalent to roughly £90 million in British currency.
Platform recommendations fuel the spread
The scope of the problem becomes evident through the use of recommendation algorithms. Researchers at Kapwing established a fresh YouTube account and monitored the content suggestions it received. Among the initial 500 recommended videos, 104 qualified as AI slop.
Roughly one-third belonged to a wider classification researchers termed “brainrot,” encompassing AI slop plus other minimal-value content designed purely for attention monetization.
This research validates what countless users have observed throughout recent months. Social feeds on YouTube, Meta’s platforms, and X increasingly overflow with repetitive, context-free, and addictive material that spreads effortlessly across geographic and linguistic boundaries.
Earlier reporting by The Guardian determined nearly 10% of YouTube’s fastest-expanding channels match the AI slop characteristics, despite the platform’s public pledges to reduce what executives call “inauthentic content.”
Global audiences drive engagement numbers
The viewer base spans continents and cultures. Spanish-language AI-focused channels attract approximately 20 million followers, representing nearly half that nation’s total population. Egypt records about 18 million followers, while the United States shows 14.5 million and Brazil claims 13.5 million, based on Kapwing’s compiled data.
Bandar Apna Dost stands out as a prominent case study. This India-based channel has accumulated 2.4 billion views featuring an anthropomorphized rhesus monkey paired with a hyper-muscular figure modeled after the Incredible Hulk. Episodes show the duo fighting demons and traveling in bizarre vehicles, including a helicopter constructed from tomatoes. Kapwing projects the channel generates up to $4.25 million yearly. The channel operator declined to answer inquiries.
Rohini Lakshané, who researches technology and digital rights issues, attributes the popularity to straightforward appeal. She highlights the ridiculous imagery, hyper-masculine themes, and absence of coherent storytelling, which permit audiences to watch without requiring prior context.
Children targeted with candy-colored content
Additional channels specifically target younger demographics. Pouty Frenchie, operating from Singapore, has registered roughly 2 billion views by chronicling a cartoon French bulldog’s adventures through candy forests and crystal sushi banquets, frequently accompanied by recorded children’s laughter. Analysts estimate annual earnings near $4 million. Cuentos Facinantes, headquartered in the United States, boasts 6.65 million subscribers, making it the most-followed operation in the research.
Some material takes concerning turns. The AI World, a Pakistan-based channel, distributes brief videos showing catastrophic flood scenarios tagged “Poor People” or “Flood Kitchen.” Multiple clips feature a soundtrack titled “Relaxing Rain, Thunder & Lightning Ambience for Sleep.” The channel has surpassed 1.3 billion views.
YouTube refuses to publish comprehensive statistics about total platform views annually or what percentage derives from AI-generated material, complicating accurate assessments. Nevertheless, researchers indicate that the pattern reveals an expanding semi-coordinated economy built around gaming recommendation algorithms.
Underground economy trades creation secrets
“There are these big swathes of people on Telegram, WhatsApp, Discord and message boards exchanging tips and ideas [and] selling courses about how to sort of make slop that will be engaging enough to earn money,” explained Max Read, a journalist tracking the phenomenon.
“They have what they call niches,” Read continued. “One that I noticed recently is AI videos of people’s pressure cookers exploding on the stove.”
Read noted many creators work from nations with robust internet infrastructure but comparatively lower wage scales.
“It’s mostly sort of middle-income countries like Ukraine, lots and lots of people in India, Kenya, Nigeria, a fair number in Brazil. You see Vietnam, too,” he said.
The enterprise lacks stability. Creator compensation systems offer little transparency, and fraudulent schemes proliferate. Some individuals selling instruction courses about viral AI content generation earn more than actual content producers.
Public awareness reaches tipping point
Meanwhile, 2025 represented a watershed moment for mainstream recognition. Meltwater, a media intelligence company, documented that mentions of “AI slop” jumped ninefold compared to 2024, with negative sentiment hitting 54% in October. SEO company Graphite calculates that AI-generated articles currently comprise over half of all English-language web content.
The terminology itself entered common vocabulary. Both Merriam-Webster and Australia’s national dictionary designated “AI slop” their Word of the Year for 2025.
Technology corporations confront mounting pressure.
“Generative AI is a tool, and like any tool it can be used to make both high- and low-quality content,” a YouTube representative stated. “We remain focused on connecting our users with high-quality content, regardless of how it was made.”
Design professionals argue that the rush toward artificial intelligence deployment frequently overlooks genuine user requirements.
“This is technology-led design, starting with the tool, and then trying to look for a problem that potentially that tool could solve,” said Kate Moran, vice president of research at Nielsen Norman Group. “That’s antithetical to how design should be done.”
Meta’s short-lived AI search test on Instagram and the underwhelming European launch of its AI video application Vibes illustrate that conflict. Internal figures reported by Business Insider revealed Vibes attracted merely 23,000 daily active users during the initial weeks.
Investment misalignment concerns researchers
For certain researchers, the challenge runs fundamentally deeper. Daniel Mügge of the University of Amsterdam suggested excessive investment flows toward applications providing minimal social benefit.
“We see that quite a bit of AI investment actually ends up in applications that make society a worse and not a better place,” he stated.
Other experts identify encouraging developments in understated tools. Moran referenced features, such as Amazon’s AI-generated product review summaries.
“Being able to give a quick qualitative summary of how people feel about that product is really valuable,” she noted.
As 2025 concludes, platforms have started introducing controls allowing users to restrict AI-generated content exposure. Whether this signals a permanent transformation or temporary accommodation remains undetermined. The undeniable reality is that AI slop has evolved from a marginal oddity to a central characteristic of today’s digital environment.
How has AI-generated content affected your online experience? Have you noticed changes in the quality of videos and posts appearing in your social media feeds? Please share your experiences and opinions in the comments below.