Online Platforms Withhold AI Content Filters as 20% of New YouTube Videos Are Called Slop
Updated
Updated · The Verge · Jun 4
Online Platforms Withhold AI Content Filters as 20% of New YouTube Videos Are Called Slop
2 articles · Updated · The Verge · Jun 4
Summary
Major platforms including YouTube, Instagram, TikTok and Spotify label some AI-made media but largely do not let users filter it out, despite growing complaints about unwanted synthetic content.
Current authentication systems rely on metadata, watermarks or detection tools that can be stripped, bypassed or trigger false positives, leaving labels unreliable at scale.
Pinterest and DeviantArt offer limited AI-content controls, but the settings are hard to find and still leave plenty of apparently AI-generated posts visible.
Platforms have already faced backlash for mislabeling human-made work, underscoring the tradeoff between broader labeling and accuracy.
The gap matters as AI content spreads widely online — one Kapwing study found more than 20% of YouTube videos shown to new users were low-quality generated material.