Snapchat announced on Friday that it will stop recommending wholly AI-generated videos in its Spotlight feed. The move aligns Snap with other tech giants like YouTube, LinkedIn, and Substack in a coordinated effort to reduce “AI slop”—low-quality, automated content flooding social platforms.
The surge in generative AI tools has made it possible to produce everything from essays to realistic videos at scale. While these tools offer utility, they have also enabled a wave of content farming that platforms are now actively fighting to preserve user trust.
Snapchat and YouTube Target Inauthentic Video
Snap’s decision to prioritize authentic, human-made content comes after acknowledging that completely AI-generated videos tend to be repetitive, low-quality, and not what Snapchat users typically want to see. However, the company is not banning AI entirely; content that is AI-enhanced or edited will still appear in user recommendations.
YouTube has taken a more aggressive approach to the financial incentives driving this trend. The video platform recently updated its policies to restrict monetization for three specific categories of videos: those that are generic, repetitive, or template-based. These styles of content are much easier to make, and quickly, with the use of generative AI tools.
This policy shift follows research from last year showing that numerous YouTube channels consisting solely of AI-generated content had amassed millions of subscribers and generated millions of dollars in revenue.
“The same technology really enables great stuff. But it also enables stuff that’s kind of content farming, and that’s the stuff that we don’t want to have.”
— Matt Halprin, YouTube trust and safety chief
LinkedIn and Substack Combat Automated Text
On the text-based side of the internet, LinkedIn and Substack are introducing tools to help users identify and report synthetic writing. LinkedIn recently launched a button allowing users to report posts or comments if they seem to be AI-generated.
The scale of the problem is significant. According to LinkedIn’s chief product officer Hari Srinivasan, who wrote on the platform that AI slop is a top priority for all of us,
the platform has blocked billions of attempts to post AI-generated comments in just the last few months, catching hundreds of thousands of automated attempts daily.
Substack is tackling the issue by providing readers with a tool to detect AI-generated writing on the newsletter platform. The urgency is underscored by research cited by Substack co-founder and CEO Chris Best, which found that up to 40% of writing on social media is now fake or AI-generated.
“Platforms that reward fakeness will create a race to the bottom.”
— Chris Best, co-founder and CEO of Substack
The Tension Between AI Utility and Content Quality
A common thread across these platforms is the attempt to distinguish between AI as a tool for refinement and AI as a replacement for human creativity. LinkedIn specifies that users who use AI tools to “refinetheir posts should not get caught up in efforts to combat AI slop. However, LinkedIn is removing an automated prompt that would show up for users writing a post offering to
enhance” it through the use of AI. It will go back to a simple proofreading tool.
Research indicates that as users encounter more AI-generated content in their feeds, they become less likely to believe that any content they see online is real. Chris Best further noted that It's getting harder to tell what's real on the internet.
By removing the financial incentive on YouTube and the algorithmic reach on Snapchat, these companies are attempting to break the cycle of content farming. The success of these measures will likely depend on whether they can effectively filter out wholly AI-generated work without stifling the “enhanced” content that still relies on human direction.
Worth a look
