A Viral Image And A Personal Breaking Point
Social media feeds are being flooded with low-quality AI-generated images and videos, a trend that is drawing criticism from users, scrutiny from platforms, and new questions about how content is created, promoted, and trusted online.
Théodore, a 20-year-old student from Paris, said the moment that pushed him to act came when he saw a viral Facebook image of two emaciated South Asian children with thick beards, one missing hands and a foot, sitting in the rain with a birthday cake while one held a sign asking for likes. The image showed multiple signs of being AI-generated, yet it collected nearly one million likes and heart emojis. He said the reaction to the post, rather than the image itself, made him start an account on X called “Insane AI Slop” to highlight and mock content he believed was misleading people. Submissions from other users soon followed, and the account grew to more than 133,000 followers.
Théodore said common themes appeared in the material people sent him, including religion, the military, and images of poor children doing emotionally charged or impressive things. He said posts showing children in poorer countries making elaborate objects or performing feats tend to perform well because people find them wholesome, which he said encourages creators to produce more of the same.
Platforms Lean Further Into AI Content
The spread of what Théodore describes as “AI slop,” meaning quickly produced and unconvincing fake images and videos, has accelerated as technology companies expand their AI tools. Meta chief executive Mark Zuckerberg said in October that social media had entered a third phase centred on AI, after earlier phases focused on content from friends and then from creators. He told shareholders that AI would make it easier to create and remix content, leading to a much larger volume of media.
Meta, which operates Facebook, Instagram, and Threads, has introduced image and video generators and more powerful filters across its platforms. When asked for comment, the company pointed to a January earnings call in which Zuckerberg said the firm was leaning further into AI and spoke about an expected rise in new media formats that are more immersive and interactive because of advances in the technology.
YouTube chief executive Neal Mohan wrote in a 2026 outlook blog post that in December more than one million channels used the platform’s AI tools to create content. He compared the shift to earlier changes brought by synthesizers, Photoshop, and CGI, and said AI would benefit creators who adopt it. He also said his team is working on ways to find and remove “low quality, repetitive content,” while adding that the company would not make judgments about what should or should not be allowed to flourish, noting that formats such as ASMR and live game streaming were once niche before becoming mainstream.
Data On The Scale Of Low-Quality AI Video
Research from AI company Kapwing said 20% of content shown to a newly opened YouTube account is now “low-quality AI video.” The firm said short-form video is a particular hotspot, with 104 of the first 500 YouTube Shorts shown to a new account falling into that category.
Kapwing said the AI-focused channel with the most views is India’s Bandar Apna Dost, which has 2.07 billion views and is estimated to earn about $4m, or £2.9m, a year. The figures highlight the role of the creator economy, where engagement and views translate into revenue, and where algorithms and audience behaviour can amplify such material.
Backlash, Moderation, And Platform Responses
A backlash has developed under many viral AI videos, with comment sections often filled with criticism. Théodore said he used his account to contact YouTube moderators about AI cartoons that he viewed as disturbing, including clips with titles such as “Mum cat saves kitten from deadly belly parasites,” which showed graphic scenes, and another in which a woman eats a parasite, turns into a monster, and is later healed by Jesus. YouTube removed the channels involved and said the content violated its community guidelines. The company said it is focused on connecting users with high-quality content regardless of how it is made and is working to reduce the spread of low-quality AI material.
Other platforms have also faced complaints. Pinterest, which is known for recipes and interior design ideas, introduced an opt-out system for AI-generated content after user frustration, though the system relies on detection or user disclosure, and continuing complaints suggest it has limits.
Across platforms such as TikTok, Threads, Instagram, and X, posts criticising AI-generated material often attract more likes than the original content. One example cited was a video showing a snowboarder rescuing a wolf from a bear, which received 932 likes, while a comment calling out the clip as AI received 2,400 likes. Despite that, all engagement still feeds platform algorithms that are designed to keep users scrolling.
How Users Judge And Verify What They See
Emily Thorson, an associate professor at Syracuse University who studies politics, misinformation, and misperceptions, said the impact of AI-generated content depends on what people are using a platform for. She said users seeking entertainment may judge content only on whether it is enjoyable, while those looking to learn or connect with communities may see AI material as more problematic.
She added that reactions also depend on how content is presented. If it is clearly framed as a joke, users may accept it as such. When it is designed to deceive, it can provoke anger. One example cited was a realistic AI-generated video of a leopard hunt, where some viewers asked in the comments which documentary it came from and asked for proof that it was not AI.
Alessandro Galeazzi, a researcher at the University of Padova who studies social media behaviour, said verifying whether a video is AI-generated requires mental effort and that over time people may stop checking. He said a large volume of low-quality AI content could further reduce attention spans. He distinguished between deceptive material and more obviously artificial clips, such as animals in unrealistic settings, but said even the latter could contribute to what he called a “brain rot” effect, where people consume content they know is unlikely to be real or meaningful.
Cases With Wider Consequences
Beyond low-quality entertainment, some AI-generated content has led to more serious concerns. Companies linked to Elon Musk, including xAI and the social platform X, changed their rules after the chatbot Grok was used to create images that digitally undressed women and children. After a US attack on Venezuela, fake videos circulated showing people thanking the United States, which analysts said could shape public opinion and create a misleading impression of support.
Dr Manny Ahmed, chief executive of OpenOrigins, a company that distinguishes between AI and real images, said people can no longer rely on visual inspection to know what is real. He said that instead of focusing only on detecting fake content, there is a need for systems that allow genuine material to prove its origin.
Major platforms including Meta and X have reduced moderation teams and moved toward approaches that rely more on users to label content as fake or misleading. That shift raises questions about whether a new platform could gain ground by promising a stricter approach, although researchers note that detection is becoming harder and that deciding what counts as low-quality content is subjective.
An Ongoing Flood And A Changed Routine
Some observers point to the rise of BeReal during the pandemic as an example of how a different approach can influence larger platforms, even if it does not overtake them. Whether a similar response could emerge around AI-generated content remains uncertain.
Théodore said he now posts less often and has largely accepted that the volume of AI-generated material is unlikely to decline. He said he is not opposed to AI itself, but to what he described as the online pollution of low-effort content made for quick views and engagement.
Featured image credits: Freepik
For more stories like it, click the +Follow button at the top of this page to follow us.