LinkedIn’s “Seems like AI slop” button has been clicked more than a million times since it launched on 31 July, and posts the platform classifies as AI slop are now getting around 40% fewer views than they did a few weeks ago.
Chief product officer Hari Srinivasan gave both figures in a LinkedIn post last week. The button sits in the three dots menu on any post and lets members flag content they believe was written by a machine. LinkedIn launched it alongside a set of changes aimed at the volume of automated content moving through the feed.
What LinkedIn changed
The company added new classifiers for identifying AI-generated posts and removed its “enhance your post” tool, which rewrote user drafts with AI. In its place is a proofreading feature that leaves the writer’s own wording intact. LinkedIn also reinforced profile and page verification.
The reports now feed back to the person who posted. Members whose content attracts enough flags see a message telling them that “Some members told us this post seems like AI.”
“We approached this assuming good intent; I know I’m increasingly conscious on how to not sound like AI & the goal is to provide helpful feedback,” Srinivasan said.
LinkedIn says it blocks hundreds of thousands of automated spam comments a day and has prevented billions of automation attempts over the past few months. Earlier this year the company said it would act against comments produced at scale “with little or no human involvement.”
Where the slop estimate comes from
The button arrived weeks after AI detector Pangram published research finding that 41% of longform LinkedIn posts were flagged as fully AI generated. 404 Media reported the study as part of wider reporting on AI spam across LinkedIn and X. Fortune, citing the same research, put the figure at 40%.
Srinivasan described the definition as a moving target rather than a fixed rule.
“Slop is hard to define and the definition changes; this lets us tune our models and make better feeds,” he said.
What it means for iGaming marketing
LinkedIn is the primary professional channel for iGaming B2B. Supplier marketing, affiliate business development, recruitment and event promotion all run through it, and a large share of that output is now produced with AI assistance. A 40% reach cut applies to those posts on the same terms as any other.
Two common practices are directly affected. The first is scaled posting, where an in-house team or agency schedules AI-drafted updates across a company page and several employee profiles at once. The second is automated commenting on other people’s posts to build visibility, which LinkedIn has been blocking since earlier this year.
Conference season raises the volume. Supplier and operator teams post heavily around events such as SBC Summit, and much of that content follows a template: an announcement, a booth number, a photo, a paragraph of praise for the panel. That is the kind of writing a classifier trained on AI text is most likely to catch, whether or not a machine wrote it.
Paid channels are moving the same way. Google has started attaching AI disclosure labels to iGaming ads, putting a visible marker on creative that used generative tools.
What operators and suppliers do not know yet
LinkedIn has not published the thresholds its classifiers use, or said whether member reports feed directly into ranking or only into training the models. A company that loses reach will not be told which posts triggered it or by how much. The flag message on an individual post is the only signal.
That leaves marketing teams working without a measurable rule. The change that matters commercially is that the cost of AI-written content on LinkedIn is no longer reputational alone. It now shows up in impressions, on a channel where most iGaming B2B deals start with a connection request.
Source: LinkedIn
