AI-generated posts are no longer just a novelty on social platforms. They are becoming a volume problem, and Snap and LinkedIn are now responding with tools and rules aimed at limiting low-quality material in recommendation feeds.
The shift is clearest on Snapchat's Spotlight, where wholly AI-generated videos are no longer eligible for recommendation as of this month. LinkedIn, meanwhile, has added a dedicated way for users to report "AI slop" as its feed faces a wave of AI-generated junk.
Snap Draws A Line Around Spotlight
Snap says it wants Spotlight to stay focused on "authentic creativity from real people" as "low-quality, repetitive, AI-generated content" spreads across the internet. That position is now reflected in how Spotlight recommendations work.
Wholly AI-generated videos are no longer eligible for recommendation as of this month. That does not mean every use of AI is banned from Snapchat's short-form video surface. Videos edited with Snapchat's own AI tools are still allowed, but they will carry transparency labels.
This distinction matters. Snap is separating content that is entirely generated by AI from content that uses AI as part of an editing process. In practice, the company is leaving room for creator tools while limiting the visibility of videos that do not come from human-made footage or human-led production.
The move also arrives as Spotlight participation has been growing. Spotlight contributors have grown by more than 120 percent, though Snap has not said how much AI content contributed to that increase. Without that breakdown, the scale of AI's role inside Spotlight remains unclear.
The Problem Is Bigger Than Snapchat
Snap is not acting in isolation. The same pressure is visible across major content platforms, especially where algorithmic recommendations can quickly amplify large amounts of similar material.
A Kapwing study found that 21 percent of YouTube Shorts are already AI-generated. The source article also notes that Instagram and Facebook likely face similar numbers. Those examples point to a broader pattern: short-form video feeds are especially exposed to AI content because they reward volume, speed, and repeatable formats.
That does not automatically make every AI-generated video low quality. But the concern raised by Snap is about repetition and feed quality. If users repeatedly encounter similar clips that feel automated or derivative, recommendation systems can start to feel less useful, less surprising, and less connected to real people.
OpenAI's social video app Sora flopped, according to the source article. That detail adds another signal: AI video may be technically easy to generate, but that does not guarantee a social experience people want to keep using.
LinkedIn Turns To User Reporting
LinkedIn is facing its own version of the issue. The platform is described in the source article as drowning in AI-generated junk, and it has rolled out a dedicated "AI slop" reporting button.
That approach depends heavily on users. A reporting button can give LinkedIn a clearer signal when posts seem low quality or AI-generated, but it also creates new problems. The source article notes that the button is vulnerable to misuse and requires users to know how to spot "AI slop" properly.
Those are meaningful limits. A user may report content because it seems generic, because they dislike it, or because they wrongly assume it was created by AI. At the same time, users may miss AI-generated content that is more polished or harder to identify. The quality of the reporting system depends on how reliably people can recognize the thing they are being asked to flag.
Why Transparency Labels Matter
Snap's use of transparency labels for videos edited with Snapchat's own AI tools shows one possible middle ground. Instead of treating every AI-assisted post the same way, the platform can tell viewers when AI tools were involved while still allowing the content to appear.
This creates a clearer distinction between disclosure and distribution. A transparency label gives viewers context. A recommendation rule affects whether the content is pushed into feeds. Snap is using both, but for different categories of AI involvement.
That framework may be important as more creators use AI features built directly into social apps. If platforms offer AI editing tools, they also need policies for how content made with those tools appears in feeds. Otherwise, the same platforms that encourage AI creation may end up overwhelming their own recommendation systems.
A New Moderation Challenge For Social Platforms
The emerging issue is not only whether content is AI-generated. It is whether platforms can preserve quality, originality, and trust when AI makes it easier to produce large amounts of material quickly.
Snap's answer is to remove wholly AI-generated videos from Spotlight recommendations while labeling AI-edited videos made with its own tools. LinkedIn's answer is to let users report "AI slop" directly. Both moves show that AI content is becoming a moderation and product design challenge at the same time.
The stakes are practical. Recommendation feeds depend on user confidence that what appears there is worth watching or reading. If low-quality AI content fills those feeds, platforms risk making their own core products feel repetitive and less human.
For now, Snap and LinkedIn are taking different routes. One is changing eligibility for recommendations. The other is adding a reporting mechanism. Together, they show that the next phase of AI content online will be shaped not only by generation tools, but by the platform rules that decide what gets surfaced.