GoKawiilLinkedIn builds AI system to demote AI-generated 'slop' posts
LinkedIn is deploying its own AI model, trained by a team of former journalists and social media experts, to detect and limit the reach of generic AI-written posts, according to CNN. A Pangram research report found over 40% of long-form LinkedIn posts are AI-generated, higher than on X, Medium, Reddit or Substack. LinkedIn also added a button letting users flag suspected AI slop, which was clicked more than 1 million times in its first two weeks after launching in August.
GoKawiil's interpretation of the reporting above, not reported fact.
LinkedIn Editor-in-Chief Dan Roth suggests that generic AI content threatens user engagement because repetitive posts fail to teach anything or deepen professional connections, which could push users toward other platforms. The company's move indicates a broader industry tension between offering AI writing tools and preserving the perceived authenticity that drives user trust and retention. LinkedIn's pairing of algorithmic detection with human 'content therapists' for creators suggests it sees this as both a technical and cultural problem.
- Over 40% of LinkedIn's long-form posts are AI-generated, per a Pangram report, more than X or Medium.
- LinkedIn's new flagging button was used over 1 million times in its first two weeks.
- LinkedIn is training its own AI model, aided by former journalists, to suppress distribution of low-quality AI posts.
Source: tech.slashdot.org — Posted, 2026-10-18
Published there as: “LinkedIn Fights AI Slop, Now 40% of Its Long-Form Posts”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.