Sunday, November 15, 2026

Top 5 This Week

Related Posts

41% of long LinkedIn posts are generated by AI


The essentials in 30 seconds

  • According to the Pangram study (more than a million posts analyzed), 41% long LinkedIn posts would be entirely generated by AI.
  • LinkedIn launched a button on July 30 “Seems like AI slop” to report this content and reduce its distribution.
  • At the same time, LinkedIn has removed its AI writing tool “Enhance your post”.
  • Warning: AI detectors remain fallible on texts hybrids (human + AI editing).

The social network for professionals has a credibility problem. According to the detection company Pangram, almost half of long LinkedIn posts are now entirely written by an AI. On July 30, LinkedIn reactedArtificial General Intelligence, a hypothetical AI at the level of human intelligence, capable of solving any intellectual task. by giving its members a button to report this “slop”. An admission as much as a response.

01 · The number

41%: the disturbing figure

Pangram analyzed more than a million publications (1,002,627 posts precisely) on five platforms, using its in-house detection model. Verdict for LinkedIn: 41% of “long” posts (at least 50 words) are entirely generated by AI. Worse, LinkedIn only represents a third of the posts analyzed but concentrates 62% of all AI content detected. The professional network is, by far, the most flooded.

41%long LinkedIn posts generated by AI

31%on Medium

25%on X (Twitter)

1.9%on Reddit, the most “human”

02 · The response

LinkedIn’s “AI slop” button

“AI slop is a top priority. People come to LinkedIn to connect with real people and share authentic perspectives. »Hari Srinivasan, Director of Product at LinkedIn (TechCrunch)

Concretely, a “Seems like AI slop” menu now allows you to report a post. The report, combined with automatic detection announced at 94% accuracy on generic content, reduces the distribution of the publication and discreetly alerts its author. This is community reporting, not automatic deletion.

03 · The paradox

The Microsoft contradiction

There’s an irony that’s hard to ignore. Microsoft, which owns LinkedIn, has invested tens of billions of dollars in generative AI and sells Copilot as a premium feature. However, at the very moment it is fighting “slop”, LinkedIn has withdrawn its own “Enhance your post” tool, which rewrote publications using AI, to replace it with a simple spell checker. The implicit message: discreet AI assistance, yes; automated full writing, no.

04 · Nuance

Beware of false positives

An AI detector is not an absolute truth. Pangram claims 0.01% false positives on pure AI text, a score validated by academics. But its accuracy drops to around 73% on hybrid texts, a human draft polished by a model. In other words, a real professional who has his text proofread by Claude or ChatGPT may be wrongly reported. And the performance of detectors in French remains undocumented. The detection versus evasion fight has only just begun.

05 · The French issue

What it changes for French pros

In France, hundreds of “personal branding” agencies sell the writing of LinkedIn posts, often mass-produced by AI. They are the first exposed to a visibility penalty. For freelancers and executives, the lesson is twofold: authenticity is once again becoming a rare and precious asset, but an assumed and reworked use of AI remains legitimate. The red line is generic content published en masse, without a point of view. Compare this with our investigation into AI disinformation in France and our daily uses of AI. Technical terms are explained in our AI glossary.

Sources

  • TechCrunch, “LinkedIn adds a button to report AI-generated slop” (July 30, 2026)
  • Xpert.digital, analysis of the Pangram study on AI content
  • SiliconAngle raises $9M from Pangram Labs (July 29, 2026)
  • The Register, “LinkedIn realizes its users have been bathing in AI slop”
  • Phrasly, critical review of the reliability of the Pangram detector

Frequently asked questions

Where does the figure of 41% of AI posts on LinkedIn come from?
It comes from a study by detection company Pangram, which analyzed more than a million posts across five platforms. For LinkedIn, 41% of posts of at least 50 words would be entirely generated by AI, and the network would concentrate 62% of all AI content detected.
How does LinkedIn’s anti-AI slop button work?
Launched on July 30, 2026, the “Seems like AI slop” menu allows you to report a publication. The report, combined with automatic detection announced at 94% accuracy, reduces the distribution of the post and discreetly alerts its author. There is no automatic deletion.
Are AI detectors reliable?
Partially. Pangram claims 0.01% false positives on pure AI text, but its accuracy drops to around 73% on hybrid texts mixing human writing and AI editing. A human text proofread by a model can therefore be wrongly reported, and the performance in French remains undocumented.
Why did LinkedIn remove its AI writing tool?
LinkedIn removed “Enhance your post”, which rewrote posts using AI, and replaced it with a simple proofreader. The message is that discreet assistance remains tolerated, but not full automated writing, despite the massive investment of Microsoft, its parent company, in AI.
Is using AI to write on LinkedIn risky?
Posting generic content mass-produced by AI now exposes you to reduced visibility and reporting. On the other hand, an assumed and reworked use of AI, with a real point of view, remains legitimate. The red line is “slop” without personal added value.
Which platforms are most affected by AI content?
According to Pangram, LinkedIn comes out on top with 41% of long AI posts, ahead of Medium (31%) and X (25%). Substack (10%) and especially Reddit (1.9%) remain the most “humane”, thanks to stronger editorial or community moderation.

Shredder Smith
Shredder Smith
Shredder Smith is the lead curator and digital persona behind topaitools4you.com, an AI directory dedicated to "shredding" through industry hype to identify high-utility software for everyday users. Smith positions himself as a blunt, no-nonsense reviewer who vets thousands of emerging applications to filter out overpriced "wrappers" in favor of tools that offer genuine ROI and practical productivity. The site serves as a watchdog for the AI gold rush, providing categorized rankings and transparent reviews designed to help small businesses and creators navigate the crowded tech landscape without wasting money on low-value tools.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Popular Articles