Behind “AI slop”: AI-generated viral video production in China
Authors ADM+S Centre
Date 27 August 2026
New research into the emerging AI “slop” industry reveals a new kind of content worker, focused on speed, scale and making content that algorithms reward.
Some creators reported earning between AUD $2,700 and $7,000 a month, while a typical creator could produce more than 50 videos a week. Successful clips could attract more than 100,000 views, while some formats, including bizarre combinations such as “pet dance motorcycle” videos, have accumulated tens of millions of views.
The research, published in AI & Society, draws on ethnographic research conducted in China between 2023 and 2025, including rare access to creators who generally keep their identities and working arrangements hidden despite producing highly visible content.
Researchers say that AI slop offers a glimpse into how AI could fundamentally change creative work.
From content creators to “post-creators” or “operators”
While traditional influencers build audiences around recognisable personas with their success depending their style or relationship with followers, AI slop creators operate differently.
Generative AI allows them to create large volumes of content, test different formats and quickly abandon those that fail to attract attention.
The capacity of AI tools to generate large volumes of content at low marginal cost enables creators to treat their cultural products as disposable experiments.
The researchers describe this as a form of “operational mastery” knowing how platforms, algorithms and production systems work and using that knowledge to maximise visibility.
In this model, creativity is less about a single person having an original idea and more about building a system capable of repeatedly producing something an algorithm will reward.
The research found AI slop creators use a range of strategies to turn virality into something that can be systematically achieved.
They benchmark content that has already gone viral, develop formulas for successful formats, test ideas before mass production and operate multiple accounts across platforms.
What happens to creativity?
Rather than a creator’s experiences, personality or identity shaping the work, creative decisions can be translated into standardised prompts, templates and production processes.
The researchers caution against dismissing AI slop simply as low-quality or annoying content.
“Rather than dismissing AI slop simply as low-quality or ‘brainrot’ content, we see it as a window into how generative AI is reshaping creative labour on platforms,” Tang said. “What is particularly interesting is that creativity doesn’t disappear, but shifts towards knowing how to test, experiment, and optimise content for platform algorithms at speed and scale.”
The content may be designed to be forgotten, but the production system behind it could tell us a lot about where creative work is heading.
Authors: Jiaru Tang, Xiaoting Yu, Patrik Wikström and Kylie Pappalardo
Read the full paper Behind the “AI slop”: exploring the AI-generated viral short video production in China
