LinkedIn is increasing efforts to limit low-quality AI-generated posts after a study revealed that over 40% of its long-form content came from artificial intelligence. CNN states that the Microsoft-owned professional network now views generic AI writing as an issue because it makes the site repetitive and less valuable.
A July study by AI-content research company Pangram showed that more than 40% of long-form posts on LinkedIn were AI-generated. This compared with about 30% on X and Medium, with lower amounts on Reddit and Substack.
LinkedIn Says Generic AI Posts All Start to Sound the Same
Dan Roth, LinkedIn Editor-in-Chief, stated that poor AI content presents a business challenge because many updates start sounding identical. When visitors repeatedly see generic text instead of educational content or genuine professional insight, LinkedIn faces the danger of users leaving for other sites. The platform previously provided AI utilities for editing or rewriting posts, though it recently reduced those capabilities.
Users Have Flagged AI Slop More Than 1 Million Times
During August, LinkedIn launched a button enabling users to flag posts that look like AI slop. LinkedIn reported that people clicked this feature more than 1 million times in the initial two weeks. The firm currently uses its own artificial intelligence tools to spot low-quality AI material and limit the reach of those posts. When LinkedIn flags a submission as AI slop, the system reduces the chances of it showing up outside the writer's immediate network.
Former Journalists Are Training LinkedIn's AI
LinkedIn formed a group mainly composed of former journalists and social media specialists to help define useful and genuine content. These specialists assist in training LinkedIn artificial intelligence models to tell the difference between valuable material and generic AI text. The organization also maintains a creator management group which Roth referred to as content therapists for executives, subject-matter experts, and other members.
Roth mentioned that certain creators now draft updates using programs like Claude or Gemini, post them on LinkedIn, and then question why the engagement is low. His group typically responds by asking if the writer genuinely liked reading the text and whether it sounded authentic to them.
AI Slop Views Have Dropped 40%
LinkedIn reports its actions are already producing results. The network announced last month that updates to detection systems led to a 40% reduction in views for material labeled as AI slop. LinkedIn does not completely reject AI assistance. Roth pointed out that not every person excels at natural communication. Still, posts left unedited, unpersonalized, or lacking authenticity now face lower chances of reaching a wide audience on the network.
