LinkedIn announced this week that it will suppress AI-generated "slop" in algorithmic feed recommendations. Posts remain visible to a user's direct connections and followers, but will not surface to other people on the platform. Suppressed patterns include "engagement bait" (Comment YES, Tag someone) and "recycled thought leadership" — generic content that, in LinkedIn's framing, "lacks the authenticity and originality" the feed is supposed to reward. The platform says AI-assisted writing is still allowed if it adds an original idea. The platform also still offers a "Rewrite with AI" button in its post composer. By one third-party estimate, 54% of long-form posts on LinkedIn are already AI-generated.

1. Finally (LinkedIn, Cory Doctorow, Garbage Day, 404 Media)

Our feeds are suffocating in machine-generated "thought-leadership." LinkedIn is the first major platform to do anything about it.

The targets here -- engagement bait, recycled thought leadership, "it's not X, it's Y" -- are slop that made the feed worth leaving. Laura Lorenzetti, who runs editorial at LinkedIn News, framed the policy in Engadget as targeting posts that "lack the authenticity and originality" that make the feed valuable. Posts you would have scrolled past faster than you scrolled past your in-laws' wedding photos are now algorithmically demoted.

The anti-slop case has been built for over a year. Cory Doctorow's March essay was titled "No one wants to read your AI slop" and called the genre a non-consensual imposition on users. Ryan Broderick at Garbage Day defines slop as content that feels worthless to the viewer, was generated in bulk, and feels forced on us by a corporation or an algorithm. 404 Media's version: "AI slop is a brute force attack on the algorithms that control reality. The intended audience of AI slop is social media and search algorithms, not human beings." By that test, the slop already won the algorithm. LinkedIn is the first major US platform trying to take the algorithm back.

2. This Is Hypocrisy (Karissa Bell, Casey Newton)

The same company that ships AI writing tools is now suppressing posts for sounding AI-generated.

LinkedIn wrote this slop in the first place! Engadget's Karissa Bell, who broke the policy story, noted in the same piece that LinkedIn "offers a bunch of its own generative AI tools, including a big 'rewrite with AI' button in its post composer." LinkedIn is owned by Microsoft, which has organized the company around Copilot. The platform spent 2024 and 2025 normalizing AI-assisted posts — and then announced 2026 as the year of suppressing them.

It's a feedback loop, and someone called it years ago. Casey Newton at Platformer, in a 2023 piece titled "The AI is eating itself," already put LinkedIn at the center of it: "LinkedIn is using AI to stimulate tired users, while Snapchat and Instagram hope bots will talk to you when your friends don't." The platform's revenue model is built on engagement. The engagement was already low. The AI tools were supposed to fix that. They produced the slop the platform is now suppressing. The cleanest read of the policy is: the company that built the problem is now being seen to fight it.

3. This Will Catch Real Writers (ghostwriting industry, copywriters)

AI detection has false positives. The LinkedIn ghostwriting market just got told its bread and butter is suppression bait.

A clean, optimized human post and a clean, optimized AI post look identical to a detector. The AI detection industry, including Originality.ai (whose own study produced the 54% stat), has documented false-positive problems: human writing that is "too perfect" — structured, free of typos, optimized for the feed — gets flagged as AI. LinkedIn says it identifies AI patterns by combining its engineering team with its in-house editorial team to recognize what adds original perspective and what doesn't. The five-year-old marketing playbook for LinkedIn posts — short paragraphs, line breaks, a hook, a list, a question — is the same playbook the AI tools learned from. If detection works on the playbook, it works on the humans who taught the playbook.

The ghostwriting market sees the storm coming. The Windmill Growth state-of-the-industry report puts the LinkedIn ghostwriting market at three times its 2024 size — the fastest-growing segment. That is exactly the use case LinkedIn says it permits in principle but is suppressing in practice. The 15-20% of funded startup founders paying $1,500-$3,000 a month for an account ghostwriter are now hoping their detector-evading style holds up.

Where This Lands

LinkedIn's policy is the first time a major US-owned platform has said out loud that it has too much AI content and will algorithmically punish it. The platform shipped most of the tools that produced the problem. The policy will probably flag humans who built modern LinkedIn-style writing in the first place. And the platform doing this is owned by the company that has bet the most on AI in productivity. All four things are true at once.

Sources