LinkedIn Tightens Controls on AI-Generated Low-Quality Content

Summary

LinkedIn has launched a multi-layered defense against polished but hollow AI-generated posts, following estimates that up to 41% of long-form content on the platform may be entirely machine-written. As part of this initiative, the company has also disabled its own AI rewriting tool.

Why it matters

LinkedIn's entire value proposition is built on genuine professional expertise and peer-to-peer knowledge exchange. Consequently, the flood of generic, low-effort AI content directly erodes the trust that makes the platform functional. This move also marks a significant turning point in the industry, as LinkedIn has drawn a clear line between using AI as a writing assistant and using it as a content substitute. This distinction will likely shape how other networks address the same issue.

Highlights

• According to Pangram, approximately 41% of long-form posts and 30% of short posts on LinkedIn between April and June 2026 were likely entirely generated by AI.
• LinkedIn removed its 'Improve your post' AI rewriting feature. This has been replaced by a more narrowly scoped grammar-checking tool that preserves the author's voice.
• Over one million users have used the new 'Looks like AI-generated content' reporting button. LinkedIn claims that flagged content experiences an approximately 40% drop in impressions.

Critical take

The two key figures in the headline—the 41% AI content share and the 40% drop in impressions—rely on detection tools and self-reported metrics. The article itself acknowledges that these tools can misclassify human-AI hybrid texts. Therefore, the true scale of the problem and the effectiveness of LinkedIn's response have not yet been verified by independent audits.