On July 30, 2026, TechCrunch's Sarah Perez reported that LinkedIn is rolling out a button letting users flag posts that "seem like AI slop." It sounds like a minor feature update. It is actually the platform sending one of its clearest signals yet that generic, low-effort AI content will increasingly lose visibility, and it should change how every business leader and marketer thinks about using AI to write.
According to TechCrunch, LinkedIn Chief Product Officer Hari Srinivasan called AI slop "a top priority" for the company. This is not a stance against AI-assisted writing itself. It is a stance against content that reads like nobody actually wrote it. That distinction matters enormously for anyone using AI tools to draft posts, articles, or updates for a professional audience.
What LinkedIn Is Actually Building
The reporting button is the visible piece, but the more consequential changes are happening underneath it. TechCrunch and The Verge both report that LinkedIn is building classifiers designed to identify AI slop and other low-quality content. Content flagged this way can see reduced suggested recommendations outside a user's existing network, meaning less reach for posts the system judges to be generic filler. LinkedIn hasn't detailed exactly how the classifiers weigh different signals, so the fair reading is that this is a quality filter in development, not a fully mapped scoring system.
LinkedIn is also testing a private signal in creator dashboards that flags when other users feel a post comes off as inauthentic or heavy on AI. That is a direct, if quiet, form of feedback: not a public shaming mechanism, but a nudge that your audience is noticing.
Perhaps the most telling change is that LinkedIn is retiring its own "enhance your post" feature, the tool that rewrote user drafts into more polished-sounding copy, and replacing it with a proofreading function built to fix words without changing voice. A platform walking back its own AI rewriting tool because it made writing sound less authentic is a strong data point about where the market is heading.
Part of a Bigger Shift
LinkedIn is not acting alone. TechCrunch notes that Substack has partnered with Pangram, a company focused on AI-content detection, to identify AI-written content on its platform. The direction across platforms is consistent. Authenticity signals are becoming standard infrastructure, not a niche feature.
For business leaders, the lesson is not "stop using AI to write." It's that detection and quality tools are improving fast enough that a policy of "we don't use AI" or "we do use AI" is no longer the right frame. The frame that matters is whether the output sounds like a real person with real judgment stands behind it.
Why This Is a Voice Problem, Not an AI Problem
AI is the instrument. A generic, unowned piece of writing is what gets flagged, whether a human or a model produced it. Plenty of AI-assisted content is edited, specific, and clearly written by someone with a point of view. Plenty of human-written content is also generic, hedge-everything filler. AI simply makes it possible to produce generic content at unprecedented scale, which is why platforms are investing in better quality signals.
LinkedIn's public messaging suggests the focus is less on whether AI was used and more on whether the resulting content adds original perspective, expertise, or context. What LinkedIn is really introducing is an implicit voice standard. The platform isn't asking whether AI helped you write. It's asking whether anyone can tell there is a real person with real expertise behind what was published.
That means AI-generated does not equal AI-assisted. Content that starts with a model but ends with a specific point of view, real examples, and a human editor in charge of the final word is a different category from content that is generated and posted with no ownership added. The real fix isn't a stricter AI usage policy. It's a voice standard: a clear point of view, specific examples instead of generalities, and edits that sound like the actual person or brand behind the post rather than a smoothed-over average of everyone else's LinkedIn content.
What Leaders and Marketers Should Do Now
The practical response is not to panic about detection tools. It's to raise the bar on what gets published under your name or your company's name, regardless of how it was drafted.
- Read every AI-assisted post out loud before publishing. If it doesn't sound like a specific person with a specific view, rewrite it.
- Replace generic claims and platitudes with one concrete detail, number, or example unique to your situation.
- Stop using any "enhance" or auto-polish feature that rewrites your voice into something smoother and blanker.
- Treat a private inauthenticity flag, if your platform offers one, as real feedback worth acting on, not an inconvenience to ignore.
If this post has you tightening the line between AI assistance and generic output, the Prompt Engineering for ChatGPT course is a practical way to get better at directing the tool without surrendering your voice. It teaches prompt structure, context, and iteration so AI drafts become raw material for sharper human judgment.*
The Takeaway
LinkedIn's slop button is a small feature with a large implication: platforms are increasingly using authenticity signals when deciding what content to recommend, not just tracking whether AI was involved. The businesses and individuals who come out ahead won't be the ones avoiding AI tools. They'll be the ones using AI to draft faster while keeping a clear, specific, recognizably human voice in charge of the final word.