How LinkedIn Is Fighting ‘AI Slop’ in Its Feed

Company leaders have described this trend as “AI slop”—posts that appear professional but are generic, repetitive, or written primarily to trigger engagement rather than contribute useful discussion.

Rather than banning AI writing tools, LinkedIn’s policy focuses on quality: content is acceptable if it reflects the user’s own voice, expertise, and perspective, even if AI helped draft or edit it.

The Types of Content LinkedIn Is Targeting

The new detection and ranking systems focus on identifying posts and comments that appear formulaic or inauthentic. Examples reported in coverage of the initiative include:

• Generic posts that contain broad motivational statements but little real insight
• Recycled “thought leadership” that repeats common ideas without new perspective
• Engagement bait designed mainly to provoke likes or comments
• Bot‑style or templated comments posted across many threads
• Formulaic writing patterns often associated with automated AI output

Both original posts and comment activity are included in the crackdown. LinkedIn specifically noted that automated or AI‑generated comments are also being targeted, since they can artificially inflate engagement or crowd out genuine discussion.

How the “AI Solving AI” Detection System Works

LinkedIn’s approach relies on AI systems analyzing patterns in text rather than simply detecting whether AI tools were used. The models evaluate signals that often appear in low‑effort AI output, such as:

• Repetitive phrasing or predictable structure
• Generic language with little personal context
• Engagement‑bait style prompts
• Lack of original expertise or perspective

By examining these patterns, the system attempts to determine whether a post provides meaningful professional value before recommending it more widely in the feed.

In early testing, LinkedIn reported that its system could correctly identify generic AI‑generated content about 94% of the time, highlighting both the scale of the problem and the company’s push toward automated moderation.

What Happens When Content Is Flagged

Importantly, LinkedIn is not automatically removing most flagged posts.

Instead, the platform primarily adjusts how widely the content is distributed. Posts or comments identified as low‑quality are less likely to appear in recommendation feeds, meaning they will reach fewer users beyond the creator’s immediate network.

This ranking‑based approach allows LinkedIn to discourage spam‑like AI content without punishing legitimate AI‑assisted writing or triggering mass content removal.

AI Assistance Is Still Allowed

LinkedIn has been clear that using AI tools to help write posts is not inherently a problem. The company’s stance is that AI can assist with drafting, editing, or organizing ideas, but the final content should reflect the author’s own thinking and experience.

Posts that include concrete examples, personal insights, or professional expertise are far less likely to be flagged—even if AI played a role in producing them.

What This Means for LinkedIn Creators

For professionals who rely on LinkedIn for networking, recruiting, or thought leadership, the shift signals a change in how content is evaluated. The algorithm is moving away from simply amplifying posts that generate engagement and toward prioritizing signal over noise.

In practice, that means content built around real experiences, specific insights, and authentic perspective is more likely to be surfaced—while generic AI‑generated filler is increasingly pushed to the margins of the feed.

As generative AI tools become more common, LinkedIn’s experiment with “AI solving AI” reflects a broader challenge facing online platforms: preserving useful conversation while managing the scale of automated content production.