40% Fake Writing Claim Pushes Snapchat to Ditch AI Slop

Up to 40% of writing on social media may now be fake or AI-generated, according to research cited by Substack chief executive Chris Best — and that figure explains why Snapchat is no longer treating AI slop as a novelty problem.

Snap, Snapchat’s parent company, said on Friday it will stop recommending “wholly AI-generated videos” in its Spotlight feed and instead favor “authentic, human-made content,” according to BBC Tech. The move puts Snapchat alongside YouTube, LinkedIn, and Substack, which have all announced anti-AI-slop measures over the past two weeks.

40% Fake Writing Claims Turn Snapchat’s Feed Change Into a Trust Test

Snapchat’s policy shift is narrow on paper. It does not ban AI from the platform. It does not reject AI-assisted creation. It targets a specific category: fully AI-generated video in Spotlight recommendations.

That distinction matters. Snap offers its own AI tools to alter content, and the company said AI “enhanced or edited” posts can still appear in recommendations. The line is not “AI bad, human good.” The line is whether the feed becomes dominated by material with no meaningful human authorship.

Snap’s own language is blunt. Entirely AI-generated content is typically “low-quality”, “repetitive”, and not what Snapchat users want to see, according to the BBC report. That is the commercial risk beneath the moderation language: if users start assuming feeds are synthetic by default, the feed loses value.

Substack’s Best framed the issue more starkly:

“It’s getting harder to tell what’s real on the internet.”

He added another warning:

“Platforms that reward fakeness will create a race to the bottom.”

MLXIO analysis: Snapchat’s move signals that major platforms are now defending trust as a product feature. The fight is not only about removing bad posts. It is about preserving the reason users scroll, creators publish, and advertisers pay for attention in the first place.

Four Platforms, Four Different Lines Around AI Slop

The latest moves show that AI content enforcement is becoming platform-specific. Snapchat, YouTube, LinkedIn, and Substack are not solving the same problem.

LinkedIn’s numbers show why the platform is acting. Chief product officer Hari Srinivasan wrote that “AI slop is a top priority for all of us.” He said LinkedIn has “blocked billions” of attempts to post AI-generated comments in just the last couple of months.

“Every day we are now catching hundreds of thousands of automated comment attempts,” Srinivasan said.

LinkedIn is also removing an automated prompt that offered to “enhance” posts with AI and returning to a simple proofreading tool. That is a quiet but meaningful retreat. The platform is not banning AI-assisted writing, but it is pulling back from nudging users toward generic machine-polished posts.

YouTube is attacking the money side. The Google-owned video platform updated monetization rules so creators cannot earn from content it classifies as “inauthentic content” — specifically generic, repetitive, or template-based videos. Research from last year found scores of YouTube channels made solely from AI-generated content, many with millions of subscribers and some making millions of dollars in revenue, according to the BBC report.

Billions Blocked on LinkedIn Show the Scale of the Moderation Problem

The strongest data point in the BBC report is not Snapchat’s policy language. It is LinkedIn’s enforcement volume: billions of blocked AI-generated comment attempts in a few months, plus hundreds of thousands caught every day.

That reframes AI slop as a scale problem. The supply can expand quickly because generative AI tools have become easier to use and can produce writing, images, and realistic video. The defense side is slower. Platforms need detection systems, user reporting, policy definitions, appeals, and enforcement teams.

MLXIO analysis: the asymmetry is the core business issue. AI-generated posts can be produced in huge volumes. Trust decisions still require context. A comment may be spam, assistance, parody, translation, or a legitimate user polishing a thought. A video may be fully synthetic, AI-edited, or human-shot with AI effects. That gray zone is where enforcement gets expensive and politically sensitive.

YouTube trust and safety chief Matt Halprin captured the tension:

“The same technology really enables great stuff.”

But he drew the line at scale-farmed output:

“But it also enables stuff that’s kind of content farming, and that’s the stuff that we don’t want to have.”

That is the operating doctrine now emerging across platforms: AI is acceptable when it supports expression, but not when it floods feeds with cheap imitation.

AI-Assisted Is Still Allowed — Fully Synthetic Is Losing Privilege

None of the platforms in the BBC report are imposing a blanket AI ban. That is the practical lesson for creators.

Snapchat still allows AI enhanced or edited content. LinkedIn says users who use AI tools to “refine” posts should not be swept up in anti-slop enforcement. YouTube says AI tools can help people with content, even as it cuts off monetization for low-quality “inauthentic content.”

The hardest cases will sit between those poles.

Creators may use AI for editing, polishing, effects, or production support. Platforms must decide whether the final post reflects a human point of view or just a template. Users want authenticity but may still enjoy AI filters and synthetic entertainment when the context is clear.

That creates an enforcement problem no label can fully solve. A simple “AI-generated” tag may be useful, but it does not answer whether the content is deceptive, low-effort, monetization abuse, or legitimate AI-assisted work.

For MLXIO readers tracking broader consumer-tech judgment calls, this sits in the same category of platform trust and product expectation questions we cover in pieces like July 30 Multiplayer Reveal Puts Gears of War: E-Day on Trial and A Fan-Cooled iQoo 16T Just Split the Flagship Fight. The subject is different, but the underlying test is similar: users punish products when the experience feels misaligned with what they came for.

Snapchat’s Spotlight Move Points to a New Feed Hierarchy

Snapchat’s decision is not just moderation. It is ranking policy.

By stopping fully AI-generated videos from being recommended in Spotlight, Snap is saying synthetic content may exist, but it does not deserve the same distribution as human-made posts. That is a more powerful tool than deletion. It changes incentives without turning every AI use case into a violation.

YouTube’s monetization limits follow the same logic. If generic, repetitive, or template-based videos cannot make money, their economics weaken. LinkedIn’s reporting button gives users a way to flag suspicious posts and comments. Substack’s detection tool gives readers a way to evaluate authorship.

MLXIO analysis: the next phase is likely to be layered enforcement rather than one-size-fits-all labels. The source material already points in that direction: recommendation limits at Snapchat, monetization rules at YouTube, user reporting at LinkedIn, and reader-facing detection at Substack.

The evidence to watch is straightforward:

• Confirmation: More platforms separate AI-assisted content from fully AI-generated posts in ranking and monetization rules.
• Confirmation: LinkedIn-style enforcement numbers keep rising, especially around automated comments.
• Confirmation: YouTube’s “generic, repetitive, or template-based” categories become a wider template for other feeds.
• Weakening signal: Platforms keep AI slop visible because it drives engagement, despite public trust concerns.

AI content is not going away. The fight is over whether platforms can make low-effort, deceptive, mass-produced synthetic content less rewarding than authentic work. Snapchat’s Spotlight change is one more sign that the answer will be written into recommendation systems, not just community guidelines.

The Bottom Line

• Snapchat is treating AI slop as a feed-quality and user-trust problem, not just a moderation issue.
• The policy draws a line between AI-assisted creativity and fully synthetic content with little human authorship.
• Major platforms are moving to protect authentic content as AI-generated posts become harder to detect.