LinkedIn cracked down on AI slop. What it means for your brand's thought leadership

If your LinkedIn feed has started to feel like the same post written a thousand times, you're not imagining it. And LinkedIn has noticed too. Over the summer and into the autumn of 2026, the platform rolled out a series of changes aimed squarely at AI-generated "slop": generic, templated posts that say very little and sound like nobody in particular.

For brands that rely on LinkedIn for executive thought leadership, this matters. The kind of content that used to be a cheap way to stay visible is now the kind LinkedIn is actively trying to show less of.

How much of LinkedIn is AI-written?

A lot. In July 2026, AI-detection company Pangram published an analysis of more than one million posts across LinkedIn, X, Reddit, Medium, and Substack. It found that more than 40% of long-form LinkedIn posts (those over 250 words) were fully AI-generated. LinkedIn made up about a third of the posts studied but roughly two-thirds of the content flagged as AI.

Detection tools aren't perfect, and different studies use different methods. But the direction is clear, and LinkedIn's own leadership has called AI slop "a top priority."

What LinkedIn has actually changed

There's been some confusion about what LinkedIn launched, so here's what's been confirmed:

• A "Seems like AI slop" report option. Added in late July 2026, it sits in a post's three-dot menu. The author gets a private notice in their analytics, and flagged posts are shown less, much like when someone taps "not interested." It was later extended to comments.
• Reduced reach for slop. In August, LinkedIn's Chief Product Officer Hari Srinivasan said the button had been used more than a million times and that content LinkedIn classifies as slop was getting 40% fewer views. That's LinkedIn's own figure, without a published method.
• Smarter classifiers. In October 2026, LinkedIn's VP of Engineering Tim Jurka said its expanded slop classifiers are 94% precise and now apply to posts distributed beyond a member's own network.
• No more AI rewrite button. LinkedIn removed its "enhance your post" rewriting tool and, in September, launched Post Proofreader for Premium users, which suggests edits while keeping the writer's voice.

LinkedIn isn't banning AI. It's deciding that content which sounds like nobody shouldn't reach everybody.

Why this matters for brand thought leadership

Plenty of corporate LinkedIn programmes run on a simple formula: take a few talking points, generate a polished post for each executive, publish on a schedule. That approach was always a bit hollow. Now it's a measurable risk. Posts that read as generic can be reported by members, downranked by classifiers, and filtered harder once they leave your immediate network, which is exactly where new reach comes from.

The same goes for employee creator programmes that hand staff a library of pre-written posts. If hundreds of people at the same company publish near-identical copy, it's hard to imagine a feed system built to spot sameness rewarding it.

Read more: Your buyers are asking AI assistants the questions your podcast should already answer

What to do instead

1. Lead with things only you could say

Specific opinions, real examples from your work, numbers from your own business, mistakes you've made. These are hard for a model to invent and easy for readers to recognise as genuine.

2. Let people hear and see the expert

Recorded conversations are one of the most convincing proofs that a real person with real expertise is behind a post. A 60-second clip of an executive explaining something on a podcast carries a voice, a face, and a point of view that a generic text post can't. That's a big part of why podcast clips work so well as the backbone of a LinkedIn programme.

3. Use AI as a tool, not an author

AI is useful for transcripts, summaries, and first drafts of supporting copy. It's a poor substitute for the opinion itself. Have a named person shape and approve every post that goes out under their name, and be open about how you use AI where it matters. If you're unsure where the lines are, our guide on whether you need to disclose AI use covers the main rules.

4. Measure depth, not just reach

Saves, thoughtful comments, and profile visits from the right people tell you far more than impressions do. If a post reaches fewer people but sparks conversations with buyers, it's doing its job.

The bottom line

LinkedIn's crackdown doesn't make thought leadership harder for brands with something real to say. It makes it harder for brands faking it. The more human, specific, and recognisably expert your content is, the better placed you are as the feed gets stricter.

If you want to build a LinkedIn presence around real conversations with your experts, talk to Poddster about turning a podcast into a steady stream of authentic content.