LinkedIn cracks down on users posting ‘AI slop’ after previously encouraging it

Microsoft Corp.’s LinkedIn is making a concerted effort to combat “AI slop,” or the low-quality, artificial intelligence-generated content that has been increasingly dominating user’s feeds. In a blog post today, LinkedIn Chief Product Officer Hari Srinivasan revealed that the company is adding a new “seems like AI slop” button for people to report posts that appear to be AI-generated.

The new button can be found in the ellipsis menu on all LinkedIn posts now, and will be coming to user’s comments too. According to Srinivasan, it will provide a way for users to report suspect posts and also help the company to train its AI-spotting AI models to enhance the quality of content found in user’s feeds.

The company isn’t stopping there, either. In addition, Srinivasan said LinkedIn is dropping the “enhance your post” button that uses AI to tweak user’s wording. It’s believed that this button probably encouraged a lot of the AI slop now found across LinkedIn’s feeds, so it’s being replaced by an option to simply have AI proofread what users have written, while leaving their natural voice intact.

The updates reflect a growing trend among online platforms to push back against AI-generated content as human users become more disgruntled by the inauthentic, computer-generated posts that now pop up seemingly everywhere they go online. Last week, for example, the popular newsletter platform Substack added a new tool that allows readers to scan new posts they’re reading to see if they were likely written by an AI model, in partnership with a startup called Pangram Labs Inc. Then, this week, Pangram revealed that it has raised $9 million in funding to try and stem the tsunami of AI content that’s now flooding the internet.

AI-generated content is a big and growing problem, and it’s hard to find anyone who hasn’t come across some kind of AI slop on platforms such as Facebook, Reddit, Quora, Instagram, TikTok or whatever social media apps they use. It’s deeply unpopular, because so many people regard AI-generated content as unreliable drivel that betrays a total lack of creativity on behalf of the poster and reduces the overall quality of both platforms and impacts user’s experiences.

LinkedIn is a particular concern, with one 2024 study finding that just over half of all long-form posts published on the site were likely faux thought leadership posts and similar slop. It has likely gotten worse in the meantime, as Srinivasan made clear in his blog post. “AI slop is a top priority for all of us,” he wrote. “We really care about this. People come to LinkedIn to connect with real people and share their real perspectives, ideas and expertise.”

The executive said the “Seems like AI slop” button is just one of several measures being introduced to try and eliminate AI content from LinkedIn. The company is also investing in automated defenses that, ironically, use AI to block “hundreds of thousands” of automated comments each day. In addition, Srinivasan talked about the introduction of new classifiers to identify if certain posts are AI slop or some other kind of low-quality content. The idea with this is to reduce the overall amount of AI content recommendations from people outside of user’s own networks. The Seems like AI slop button will be one of several signals LinkedIn uses to inform its classifiers, Srinivasan said.

In addition, LinkedIn will also start flagging user’s dashboards privately when it thinks they’re using AI too heavily and their content seems “inauthentic.” Srinivasan said this is meant to encourage users to improve their writing in cases where they simply use AI to enhance what they’ve written, rather than target users who post entirely AI-generated posts.

Image: LinkedIn