LinkedIn Battles AI 'Slop' to Reclaim Human Connection

LinkedIn, once a bastion of professional networking, is now grappling with an influx of bland, AI-generated content that has become a frequent target of online humor. The platform has acknowledged user complaints about lengthy, generic posts devoid of human touch. As artificial intelligence makes content creation cheaper and easier, the internet has been inundated with such "slop." Platforms like YouTube, Pinterest, and Spotify have already implemented measures to curb the visibility of spammy AI-generated content, but LinkedIn has been particularly affected.

Earlier this year, users scrolling through LinkedIn could encounter a barrage of bullet-pointed lists, emojis, references to "B2B sales," and calls for follows or comments. A July report by Pangram, a research firm monitoring online AI content, revealed that over 40 percent of long-form posts on LinkedIn were AI-generated, a figure higher than X and Medium (around 30 percent) and significantly more than Reddit and Substack.

In response, LinkedIn is intensifying its efforts to combat this low-quality content. Dan Roth, LinkedIn's Editor-in-Chief, stated in an interview with CNN that while users might believe their perfectly formatted posts are effective, authenticity is key. "But what works is being human," he emphasized.

This situation presents a notable irony, given that LinkedIn is owned by Microsoft, a company actively promoting its AI tools across various sectors, including writing and coding, and facing stiff competition from entities like OpenAI and Anthropic. LinkedIn had previously offered AI-powered editing and rewriting tools but has recently reduced their availability. However, the proliferation of low-quality AI posts poses a significant business challenge for LinkedIn. Roth explained that these posts tend to sound alike, and if users encounter too many generic AI-generated pieces instead of insightful content that fosters genuine professional understanding, LinkedIn risks losing its user base to competing platforms.

To combat AI "slop," LinkedIn introduced a feature in August allowing users to flag posts as potentially AI-generated. This button received over 1 million clicks within its first two weeks. The company is now leveraging its own AI systems to identify and reduce the distribution of such content, while also guiding users on how to create more authentic posts. A dedicated team, largely composed of former journalists and social media experts, is training LinkedIn's AI model to recognize genuine, valuable content. Posts flagged by the AI as low-quality and AI-generated will have their distribution limited, typically to the user's direct connections.

Additionally, a creator management team acts as "content therapists" for executives and subject matter experts. Many have become accustomed to the ease of using AI for post creation, only to find their content underperforming. "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 shared. The team encourages users to embrace imperfection, suggesting methods like recording videos with a phone propped against a coffee mug instead of a professional studio, or posting incomplete thoughts to spark conversation.

These initiatives are yielding positive results. LinkedIn reported a 40 percent decrease in views of content classified as AI "slop" following improvements to its detection systems. Roth clarified that LinkedIn is not entirely against AI, acknowledging that not everyone is a natural communicator. However, he stressed that failing to personalize or edit AI-generated content diminishes its reach on the platform. "We just try to push people to be more human," he concluded. "That is what is required to be able to cut through the noise."