There is a special purgatory in the modern internet, and it looks like a LinkedIn feed: bullet-pointed confessions about resilience, festooned with emojis, ending always with an invitation to comment for more. The professional network has spent the year as the punchline of the internet’s longest joke, and now, as Clare Duffy reported for CNN, it has begun to fight back against the flood of machine-made sincerity that drowned it.

The numbers explain the desperation. Pangram, a research firm that tracks AI content online, found in a July report that more than 40% of long-form posts on LinkedIn were AI-generated — a higher share than the roughly 30% on X and Medium, and far above Reddit and Substack. Everywhere the machines write, but somehow they write most industriously about B2B sales.

The irony of the house

Here the story acquires its bite. LinkedIn belongs to Microsoft, the corporation that has spent these years urging humanity to write, code and think with its AI tools, locked in steep competition with OpenAI and Anthropic. LinkedIn itself once offered AI features to edit and rewrite posts, before quietly scaling them back. The empire that sold the sludge now pays to dredge the river. One is reminded of kings who taxed the plague.

Dan Roth, LinkedIn’s editor-in-chief, who leads the platform’s content strategy, delivered the verdict to CNN: “I know you think you’re doing the right thing with your perfectly bulleted list and your hook at the top, and you followed all the advice from LinkedIn advisors across the internet. But what works is being human.”

The commercial logic is naked. Generic posts all sound the same, Roth said, and if users conclude the feed teaches them nothing real about their professional connections, they will drift elsewhere. Slop is not an aesthetic problem; it is a churn problem.

A million small denunciations

In August, LinkedIn handed its citizens a small instrument of revolt: a button to flag a post that “seems like AI slop.” It was clicked more than 1 million times in its first two weeks. Now the company fights machine with machine — an AI model trained to recognize synthetic drabness, schooled by a team composed largely of former journalists and social media experts, people paid to know what an authentic post looks like. Posts the model condemns as low-quality and AI-generated will be confined to their authors’ immediate circle, unpublished in the public square.

And then there is the strangest bureau of all: a creator management team Roth describes as “content therapists,” counselors for executives and subject-matter experts who have grown dependent on the convenience of the ghost.

“A lot of times now today, they’ll run it through Claude or Gemini or wherever, they’ll post it on LinkedIn and it will perform poorly. And then they come to the team and they’re like, ‘Why did this do badly?’” Roth said. “And the best thing you can do is give it back to them and be like, ‘Did you enjoy reading this? Do you see yourself in this?”

The aesthetics of the coffee mug

The prescription borders on doctrine: embrace imperfection. Lean your phone against a coffee mug to film a video instead of hiring a studio. Post the unfinished thought that starts a conversation. The platform claims early results — a 40% decline in views of content classified as AI slop, it said last month, attributing the fall to improved detection.

None of this makes LinkedIn an enemy of the machine. “Not everyone is a born communicator,” Roth conceded. The sin is not using AI; it is failing to edit, to personalize, to leave a fingerprint. “We just try to push people to be more human,” he said. “That is what is required to be able to cut through the noise.”

A corporation now employs therapists to cure its users of the convenience it helped sell them. Whether this is reform or merely better marketing of authenticity, the feed will decide.