YouTube is now asking viewers if videos feel like AI slop

YouTube is testing a new viewer-driven feature that asks users to rate whether a video feels like low-quality AI content, adding a third layer of detection on top of its existing automated and human review systems. The move comes amid mounting evidence of an AI slop problem on the platform — one study found that around 21% of videos recommended to new accounts qualify as AI slop, while a third fall into a broader "brainrot" category. YouTube has not explained how the collected ratings will be used, and at least one observer has raised concerns that the data could end up as training material for Google's own AI video models, potentially making future AI content harder to detect.

Read the full story on Digital Trends →

Editor's Note: The real problem is that YouTube's algorithms have been amplifying low-quality content long before AI slop became a talking point — things were already out of control. On top of that, this experiment has a built-in bias risk: many English-speaking users tend to flag AI-dubbed videos as "AI slop" even when the original content was made by real creators, which could skew the data and penalize legitimate non-English creators unfairly.