LinkedIn Removes AI Writing Tool, Adds 'AI Slop' Button to Combat Generic Content Flood

LinkedIn Killed Its Writing Assistant. It Wants Your Thinking, Not Your First Draft

LinkedIn just announced a "Seems like AI slop" button that lets users flag posts that appear AI-written. The Microsoft-owned network also removed its own AI writing feature and replaced it with a tool that proofs your post while leaving your writing style alone. LinkedIn's product chief said slop is a top priority, and the company now blocks hundreds of thousands of automated comment attempts a day. The move follows Substack, which last week added a scanner from Pangram that estimates how much of a post was written by AI. It also arrives as Cloudflare reports that bots now generate 57 percent of web traffic to sites it hosts. This level of bot traffic wasn't expected to occur until 2027. When it comes to writing, AI is a double edged sword. It's fast and easy. But it's also relatively generic without a highly detailed prompt. The result is usually slop that isn't worth ready and that can undermine the very message being posted or email being sent. When you post the raw output of your AI results without thoughtfully editing the content, you're guilty of slop. Platforms Are Picking Sides LinkedIn's decision to kill its own writing assistant is telling. A platform that purposely helped people generate posts has seemingly decided that generation itself was the problem. Linkedin's replacement corrects your grammar and hands the writing back to you. The message to users is clear: bring your own thinking, and let the AI clean things up. Substack made the same decision from the other direction. Its CEO has described the right use of AI as handling everything "except the hard part," where the hard part is having an idea worth sharing. Both companies are betting that users will find real human engagement more sticky than conversing with AI- content generated mindlessly by other users.

Tired of AI Slop on Your LinkedIn Feed? Here's How the Company Is Pushing Back.

AI detection firms estimate that a substantial share of LinkedIn posts and comments now involve AI, though LinkedIn disputed their results. LinkedIn knows it has an AI problem. Last month, the professional networking platform started giving users a way to flag posts and comments they believe were created by AI, using a "seems like AI slop" reporting button, The Wall Street Journal recently reported. The move aims to slow the flood of AI-created content. LinkedIn isn't alone. Most major social media platforms have tried to curb the rise of AI-generated posts, largely by relying on automated labels. It's an approach with a spotty record, leaving some AI work unlabeled and mislabeling some human-created content as AI. For example, social media star Nikolai Savic, who has five million followers on TikTok, is well-known for his creative cooking videos. He told the Journal last month that TikTok labeled some of his videos as AI-generated, even though he painstakingly edited them himself. The move caused him to lose credibility with fans. "People already have in their minds that I'm only using AI," Savic said. "This hurt my reputation so much. And at TikTok, nobody really seems to care." The stakes may be higher for LinkedIn, which sells itself as a trusted place to find useful information, build professional relationships and uncover career opportunities for its more than one billion members. Many LinkedIn users turn to AI hoping it will help them communicate more clearly or post more often. Executives and industry observers caution that the result can be content that feels bland, overly polished or inauthentic. AI as a new challenge Brendan Gahan is the co-founder and CEO of Creator Authority, a LinkedIn influencer marketing agency. He told the Journal that "everybody uses AI to some degree." However, he advises influencers to skip posting AI-generated content because it could damage their credibility. Success on the platform ultimately depends on credibility, he said. LinkedIn has encountered content problems before, like oversharing and scam job posts. "We have been working on this domain of low-quality content for a long time," Oscar Rodriguez, LinkedIn's vice president of product, told the Journal. AI is testing the platform on a much larger scale. From April through June, 41% of LinkedIn's long-form public posts and 30% of its public comments were entirely AI-generated, according to Pangram Labs, an AI-detection startup that analyzed nearly 57,000 pieces of content. The company found that LinkedIn had a higher concentration of AI-generated long posts than X, at 29%, and Reddit, at 13%. Originality.ai, another AI-detection company, reached an even more striking conclusion. After reviewing 5,000 public LinkedIn posts in July, it determined that 81% showed more than a moderate amount of AI use. By comparison, just under half of the 5,500 public Reddit posts it reviewed met that same threshold. LinkedIn disputed the research findings to the Journal, but did not share comparable results of its own.