How LinkedIn is cracking down on generic AI-written content
What's the story
LinkedIn, the professional networking platform, is taking a stand against bland and generic AI-generated posts. The move comes after users have complained about long-winded missives that lack human touch. A report from Pangram in July revealed that over 40% of long-form posts on LinkedIn are AI-generated. This is higher than other platforms like X and Medium where the percentage hovers around 30%. The percentage is even smaller on Reddit and Substack.
Strategy
A new button allows users to flag 'AI slop'
In response to the influx of AI-generated content, LinkedIn has launched a button for users to flag posts that appear as "AI slop." The feature was used over a million times in its first two weeks.
Now, the company is leveraging its own AI systems to filter out such content and remind users how to create more authentic posts.
Team effort
Creator management team trains AI to identify authentic posts
LinkedIn has formed a creator management team, made up mostly of former journalists and social media experts.
This team helps train the platform's AI model to identify authentic posts.
If the model detects a post as low-quality or AI-generated content, it will be less likely to reach beyond that user's direct connections.
Authenticity promotion
'Content therapists' for executives and subject matter experts
The creator management team also acts as "content therapists" for executives, subject matter experts, and other LinkedIn users.
They encourage these users to embrace imperfection in their posts.
The strategy has shown early signs of success, with improvements to LinkedIn's detection systems resulting in a 40% decline in views of content classified as AI slop last month.