In a 24-hour window at the end of July 2026, two platforms that agree on almost nothing drew the exact same line against AI slop. On July 30, LinkedIn — a professional text network — added a "Seems like AI slop" button to every feed post, handing members a one-tap way to flag machine-written content and feed its suppression classifiers. On July 31, Snapchat — an ephemeral short-video entertainment app — adjusted Spotlight so wholly AI-generated videos are no longer recommended, even when disclosed. Read separately they are two product updates. Read together they are the more important story: when the two extremes of the social internet, one built on candid camera footage and one built on written professional credibility, converge on the same enforcement in the same week, the quality bar is no longer a niche policy — it is an industry standard, and it now spans both axes of content at once. The two platforms even chose opposite enforcement mechanisms — Snapchat's is a top-down, invisible algorithmic deprioritization; LinkedIn's is a bottom-up, crowdsourced human report — which together bracket the full range of how a feed can now push AI slop down. This guide is the synthesis: what each platform actually changed, why the pairing matters more than either announcement alone, the two content axes (video authenticity and text voice) you now have to clear simultaneously if you publish across platforms, why neither is a ban on AI, what genuinely clears both bars, and how to run identity-anchored content at volume across the whole spread — video and text — without tripping either mechanism.
At the end of July 2026, two platforms that agree on almost nothing moved against AI slop inside a single 24-hour window. On July 30, LinkedIn added a "Seems like AI slop" button to the three-dot menu on every feed post — a one-tap way for members to flag content that reads as machine-written. On July 31, Snapchat adjusted its Spotlight recommendation systems so that wholly AI-generated videos are no longer recommended, even when the creator discloses the AI. Two announcements, one day apart, from a professional text network and an ephemeral video app that share virtually no audience, format, or culture.
Read individually, each is a routine product update with its own dedicated write-up — the mechanics of the Snapchat change are covered in Snapchat deprioritizes fully AI-generated content in Spotlight, and the reach mechanics of the LinkedIn button in LinkedIn's "Seems like AI slop" button. This guide is deliberately the other thing: the read on the pairing itself. When the two extremes of the social internet draw the same line in the same week — and pick opposite enforcement mechanisms to do it — the story is no longer either platform's policy. It is that the quality bar has become an industry standard spanning both ends of content at once, video and text, and that changes how anyone publishing across platforms has to operate.
Start with how unalike these two are, because the difference is the whole point. Snapchat's Spotlight is a trending entertainment feed built on candid, camera-shot moments — short vertical video, an audience skewing young, a culture of the raw and the immediate. LinkedIn is a professional network whose entire value proposition is credibility — long and short written posts, a culture of career signaling, an audience of professionals and marketers. They do not compete for the same creators. They do not reward the same content. If you set out to pick two major platforms with the least in common, you would land somewhere close to this pair.
Which is exactly why their agreement is worth noticing. A single platform tightening its AI rules is a local decision — it tells you about that company's feed and its particular problem. Two platforms this far apart tightening within 24 hours of each other is a market-wide signal. It says the pressure to raise the quality floor is not coming from one product team's taste; it is coming from a condition every distribution surface now shares — feeds filling with cheap synthetic content faster than passive ranking can absorb it. When the video extreme and the text extreme both respond in the same week, the safe read is that everything in between has the same problem and is moving the same way. This is the convergence the broader survey in platforms are battling AI-generated spam maps across the rest of the industry; Snapchat and LinkedIn are the two clearest bookends of it.
The specifics matter, because the two platforms gated different things in different ways, and your response depends on which mechanism is pointed at you.
On July 31, 2026, Snapchat said it adjusted the recommendation systems behind Spotlight so that only videos made by real people — not wholly AI-generated ones — are eligible to be recommended. Its stated reasoning was direct: as low-quality, repetitive, AI-generated content becomes common across the internet, it wants Spotlight to remain a place to discover authentic creativity from real people. Two details define the change. First, disclosure does not buy a fully synthetic clip back in — even a properly labeled AI video is no longer surfaced. Second, it is not a ban: you can still post AI content, and AI used to enhance or edit real footage, including Snapchat's own creative tools, stays eligible and monetizable. What changes is recommendation weight, applied silently — there is no notice when your synthetic clip stops reaching people who do not already follow you. The event write-up is in Snapchat stops rewarding fully AI-generated Spotlight content.
On July 30, 2026, LinkedIn added a "Seems like AI slop" option to the three-dot menu on every feed post. Tapping it does not remove the post; it returns a short acknowledgement and feeds a labeled signal into LinkedIn's detection classifiers, which suppress content judged to be slop from recommendations so it stops circulating beyond the author's existing connections. The poster gets a private nudge in their analytics dashboard when their writing reads as inauthentic, and LinkedIn said the feedback helps determine how much reach a post gets outside the poster's network. It moved harder than most because it has the worst version of the problem: an analysis by Pangram Labs found roughly 41% of long-form and 30% of short-form LinkedIn posts were likely AI-generated. In the same push, LinkedIn retired its own "enhance your post" AI writer — the tool that homogenized voices — and replaced it with a proofreader that corrects without rewriting. The launch details are in LinkedIn's AI slop button.
Line the two up and the striking thing is that they chose opposite ways to enforce the same standard. Snapchat's mechanism is top-down and machine-driven: a change to what the algorithm recommends, invisible to the creator, judging content on whether a real person and real capture are behind it. LinkedIn's mechanism is bottom-up and human-driven: a button that hands the judgment to your own audience, so detection is partly crowdsourced to the readers scrolling past your post. Between them they bracket the full range of how a feed can now push AI slop down — pure algorithm at one end, pure human sniff test at the other.
The practical consequence of that range is that you cannot game your way clear of it. A paraphrasing trick or a disclosure label might fool a classifier, but Snapchat's system is looking for the presence of a real human behind the video, and LinkedIn's is a professional deciding in real time whether your post reads as filler. Neither is a keyword filter you can reverse-engineer. And both share a quiet trap: they give you almost no feedback. Snapchat will not tell you your clip stopped being recommended; LinkedIn suppression looks identical to a post that simply did not resonate. The penalty is silent on both, which means you cannot debug your way out after the fact — you have to clear the bar before you publish, not diagnose why you did not afterward.
The deeper reason the pairing matters is that Snapchat and LinkedIn gate two different axes of content, and a creator publishing across platforms has to clear both at once. Snapchat's line is about video authenticity: is there a real person and real capture anchoring this clip, or is it anonymous synthetic footage dropped in wholesale? LinkedIn's line is about text voice: does this writing carry a specific, original point of view, or is it the generic, templated register — the much-mocked "it's not X, it's Y" cadence — that reads as the average of the internet? Passing one says nothing about passing the other. A flawless human-shot video with a generic AI caption fails the text bar; a sharp, original written take rendered as a faceless synthetic clip fails the video bar.
This is where most creators get caught, because the tools they reach for solve one axis and ignore the other. A clipping or avatar-video app can help you clear the video-authenticity bar but does nothing for how your written posts read. A writing assistant can sharpen a LinkedIn post but has no bearing on whether your Spotlight or Reels clip looks human-anchored. The operator publishing across the whole spread — short video on the entertainment feeds, written posts on the professional ones — now has to satisfy two separate standards simultaneously, which is a harder problem than either platform's announcement makes it sound. The volume-and-saturation side of the same squeeze on the text axis is worked through in how saturated LinkedIn and X really are.
It would be easy to file both moves under "platforms turning against AI," and that misread leads to the wrong fix. Neither platform banned AI as a category, and both said so plainly. Snapchat still rewards AI that enhances real footage and still lets you post synthetic content; it just will not recommend the fully synthetic version. LinkedIn explicitly permits "AI-assisted" content that carries original ideas and starts real conversations; it targets the generic, low-perspective, obviously machine-written kind. Set the two side by side and the shared target is not the use of a model — it is content with no identifiable human and no original substance behind it. Snapchat calls it wholly AI-generated and anonymous; LinkedIn calls it generic slop. Same judgment, two vocabularies.
That reframing inverts the intuitive response. The instinct on hearing "two more platforms cracked down on AI" is to use less AI. That is the wrong lever, because the behavior being penalized is not AI use — it is the absence of the things that make content worth distributing. You can post anonymous synthetic filler with no AI at all and get buried; you can publish heavily AI-assisted content with a real identity and an original angle and clear both bars comfortably. The detectors, human and machine, are graded on provenance-of-a-human and on substance, not on whether a model touched the work. The same principle, traced from the demand side as slop floods every feed, is the subject of the AI slop content trend, and the monetization version of it is in YouTube's AI content policy.
Strip the two standards to their shared requirement and the same word falls out of each: identity. Snapchat's video bar is cleared by a real, recognizable person anchoring the clip — a face, a voice, real capture that AI serves rather than replaces. LinkedIn's text bar is cleared by a specific point of view a reader recognizes as unmistakably yours — writing that could only have come from your material and angle, with the generic AI tells stripped out. Both are, at root, the same demand: a consistent, identifiable human presence behind the content, on whichever axis the platform happens to gate.
The second shared requirement is per-platform nativeness. The fastest way to fail both bars at once is to produce a single asset and blast it everywhere — the identical clip cross-posted to every video feed, the identical text pasted into every written one. That copy-paste sameness is exactly the undifferentiated pattern both mechanisms are tuned to catch, and it multiplies your exposure: one generic asset fails on every surface it lands. Clearing the bars means generating content native to each destination, anchored to the same identity, with a human deciding what ships. That is straightforward to say and hard to do by hand at any real cadence — which is where the workflow question comes in.
The awkward part of the Snapchat-and-LinkedIn squeeze is that it asks you to be excellent on two opposite axes at once — human-anchored video and non-generic written voice — usually with two different toolchains that do not know about each other. Kompozy is built as one engine that generates net-new on both. On the video axis it produces persona and avatar shorts, clipped verticals, and marketing shorts with a consistent AI Influencer persona — a stable face and voice carried across every clip, which is precisely the human-anchored signal recommendation feeds like Snapchat's now reward over anonymous synthetic footage. On the text axis it writes posts, blogs, and newsletters governed by a Persona Brief and banned-word and voice filters that keep the copy out of the generic AI register a LinkedIn reader flags on sight. One system, both bars.
The neat part is how directly the two capabilities map onto the two enforcement mechanisms. Against Snapchat's invisible, algorithmic "is a real person behind this" test, Kompozy's answer is a persistent persona identity attached to every video output, so the human-anchored signal is native rather than bolted on. Against LinkedIn's human report button and its inauthentic-voice classifier, Kompozy's answer is generating from your own material with a defined voice and stripping the templated cadence, plus a review pipeline that keeps a person approving what goes out — the human-in-the-loop checkpoint both platforms are effectively demanding. And because it publishes natively to each destination rather than copy-pasting one asset, it holds the per-platform differentiation that keeps you off the sameness pattern; LinkedIn is one of the eight social platforms it publishes to directly, alongside blog and email.
The honest scope: Snapchat is not one of Kompozy's direct publishing destinations, so this is not a claim that it posts to Spotlight for you. But Snapchat is the crispest public statement of a video-authenticity standard that now governs the recommendation feeds Kompozy does publish to — TikTok, YouTube, Instagram, and the rest — so building content that clears the Snapchat line is building content that travels on those surfaces too. The value is that you are not stitching a video tool to a writing tool and hoping both stay on-brand; you are running one identity across both axes at the volume the platforms will actually distribute. For the wider strategic frame on operating across a saturated professional feed, how saturated LinkedIn and X really are picks up the text side in depth.
Snapchat and LinkedIn cracking down on AI slop within 24 hours is best read not as two stories but as one: when the two most dissimilar corners of the social internet draw the same line in the same week, the quality bar has become an industry standard, and it now spans both the video and the text ends of content. They even chose opposite mechanisms — invisible algorithm and crowdsourced human report — which together define the full range of how a feed can push slop down, and neither of which you can game. The move that clears both is not less AI; it is a real, consistent identity anchoring your video and a specific, non-generic voice carrying your text, produced natively for each platform with a human in the loop — at the volume the feeds will now actually reach.
Within 24 hours at the end of July 2026, both moved against AI slop in different ways. On July 30, LinkedIn added a "Seems like AI slop" option to the three-dot menu on every feed post, letting members flag content that looks machine-written; those reports train its classifiers, which suppress generic AI content from recommendations. On July 31, Snapchat adjusted its Spotlight recommendation systems so wholly AI-generated videos are no longer recommended, even when disclosed. Neither banned AI — both target anonymous, fully synthetic, or generic templated content while still allowing AI used to enhance real work.
Because they are about as different as two platforms get — Snapchat is an ephemeral short-video entertainment app built on candid camera footage; LinkedIn is a professional network built on written credibility. They share almost no audience, format, or culture. When two platforms this dissimilar draw the same line against AI slop in the same week, it stops looking like one company's niche policy and starts looking like an industry standard. The convergence is the signal: the quality bar now spans both the video and the text ends of content at once, not just one corner of the internet.
No, and both were explicit about that. Snapchat still lets you post AI content and still rewards AI used to enhance or edit real footage — including its own AI creative tools; it only stopped recommending wholly synthetic videos in Spotlight. LinkedIn still permits "AI-assisted" posts that carry original ideas and start real conversations; it targets generic, templated, obviously machine-written slop. In both cases the target is a behavior — anonymous synthetic volume or generic templated voice — not the use of AI itself.
They sit at opposite ends of the spectrum. Snapchat's is top-down and invisible: an algorithmic change to what Spotlight recommends, with no notice when your fully AI-generated clip stops being surfaced. LinkedIn's is bottom-up and human: a report button in your readers' hands, feeding classifiers that suppress reach outside your network, plus a private dashboard nudge when your own writing reads as inauthentic. One is a machine judging provenance; the other is a real person deciding your post smells like a bot. You cannot game either, and together they cover the full range of how a feed can push slop down.
Treat them as two separate tests you have to pass at once. For the video bar (Snapchat and the recommendation feeds like it), anchor clips to a real, consistent identity — a face and voice a viewer recognizes — and use AI to enhance real capture rather than replace it. For the text bar (LinkedIn), generate from your own material so the output carries a specific point of view, strip the recognizable AI tells, and write natively for each platform rather than pasting identical copy. Keep a human review step before anything ships, and attach a consistent persona to everything so no post reads as anonymous or generic.
At the end of July 2026, Snapchat and LinkedIn cracked down on AI slop within 24 hours of each other. On July 30, LinkedIn added a "Seems like AI slop" report button that feeds classifiers suppressing generic AI posts; on July 31, Snapchat stopped recommending wholly AI-generated videos in Spotlight, even when disclosed. Neither banned AI. The pairing matters because two maximally different platforms — one video, one professional text — drew the same line at once, signaling an industry-wide quality bar that spans both content axes.
Get started → · ← All guides · Compare Kompozy vs other tools