Snapchat is one of a growing number of tech platforms that are moving to dial down the prominence of fully AI-generated content, often labelled “AI slop", amid worries about the authenticity and quality of online content.
Snap Inc., the parent company of Snapchat, said on Friday it would stop recommending videos that were completely made by artificial intelligence on its Spotlight feed. Instead, the platform said it would focus on content created by people but would continue recommending posts that use AI as an editing or improvement tool.
The move follows similar actions by YouTube, LinkedIn and Substack over the past two weeks, as part of a wider industry push to tackle the rapid rise of AI-generated text, images and video on the internet.
Generative AI tools have made the generation of realistic content easier than ever, leading to an explosion of synthetic media on the internet. While many companies still tout AI-powered creative tools, a few are now rolling out policies to restrict low-quality or misleading material.
Snap said it’s not banning AI-generated content outright. Snapchat users will still be able to create and share content with the app’s AI features, as long as they use AI to enhance or edit existing content, rather than generate entirely new videos from scratch.
Fully AI-generated videos, the company said, tend to be repetitive, low quality and less in line with what users expect to see in Spotlight.
The company cites research showing audiences are less likely to respond positively to content that is fully AI-generated. Separately, survey findings also show that frequent exposure to synthetic media can reduce users’ trust in the authenticity of content across social media platforms.
Other platforms tighten AI rules
Substack has just launched a tool to help readers spot AI-generated writing. In announcing the feature, Co-founder and Chief Executive Chris Best said it is getting harder to tell real content from content created by AI.
Best also referenced research indicating that up to 40% of writing on social media may now be AI-generated or otherwise inauthentic, warning that platforms that reward such material risk encouraging lower-quality content.
LinkedIn has also stepped up its fight against AI-generated spam. The professional networking platform recently rolled out a feature that allows users to report posts and comments they suspect are AI-generated.
LinkedIn’s chief product officer, Hari Srinivasan, said the company has blocked billions of attempts to post AI-generated comments recently and now detects hundreds of thousands of automated comment attempts each day.
LinkedIn said it is not discouraging the responsible use of AI to improve writing. But it’s taking away an automated prompt that used to allow users to boost posts with AI and replacing it with a basic proofreading tool.
YouTube has also updated its monetisation policies. The platform said earlier this month that videos deemed generic, repetitive or template-based will not be eligible for advertising revenue under new rules targeting “inauthentic” content.
The changes come after previous studies found hundreds of YouTube channels built almost entirely on AI-generated videos, some of which have attracted millions of subscribers and are generating revenue.
In a statement on the policy update, YouTube Trust and Safety Chief Matt Halprin said AI remains a valuable creative tool but is increasingly used to produce low-quality, mass-produced content the platform wants to discourage.
Recent announcements suggest major online platforms are attempting to balance support for AI-assisted creativity with growing concerns about authenticity, user trust and the spread of repetitive automated content.
