Key Takeaways
- LinkedIn added a new feature last month: a button users can press if the content they’re seeing “seems like AI slop,” or poorly generated AI content.
- In the first two weeks of the button’s launch, more than a million people have used it.
- AI-detection startup Pangram recently classified more than 40% of LinkedIn’s long-form posts as entirely AI-generated.
LinkedIn doesn’t want to lose users to a flood of AI content.
In an effort to create a better user experience, LinkedIn chief product officer Hari Srinivasan announced in a post on the platform last month that the company would include a new feature: a “seems like AI slop” button.
AI slop refers to poorly generated AI content. LinkedIn is allowing users to flag posts that appear to be AI-generated.
The button immediately proved popular. Srinivasan wrote on LinkedIn last week that within the first two weeks of the button’s launch, more than a million people have used it.
“Overall members are now experiencing 40% less views on what we classify as AI slop from just a few weeks ago,” he wrote in the post. “Despite the progress, we know we have more to do to ensure LinkedIn remains a place where you can find real people & real perspectives.”
Avoiding activity from AI bots
LinkedIn is trying to avoid the “dead Internet theory,” or the idea that much of what users see online is no longer made by real people for real people. The theory imagines an Internet increasingly filled with bots, AI-generated posts, fake accounts and automated engagement.
In other words, machines effectively post, like, comment and boost each other, making most of the content on the Internet AI-generated.
As AI agents and bots flood online spaces, the theory has gained traction.
“There’s a lot of missing pieces of information, but a lot of the observed data suggests the same thing, which is there is more bot activity,” Rudy Yang, PitchBook’s enterprise and retail fintech analyst, recently told Fortune. “Agentic AI activity is driving a lot of the browser activity you’re seeing.”
AI content is gaining ground
Data supports the idea that the web is increasingly dominated by AI. Cloud platform Cloudflare said in April that for the first time, automated traffic had overtaken human traffic on the Internet. As of Wednesday, bots generated 58% of search requests, compared to 42% from people, per Cloudflare.
AI-written material is also showing up more often across the web. In a Pew Research Center analysis released last week, researchers examined 10,000 webpages gathered in July and found that 10% contained strong indicators of AI authorship, up from roughly 2% five years earlier.
More than one-third of webpages published since ChatGPT’s late-2022 debut showed at least some evidence of AI-generated writing, the analysis added.
LinkedIn appears especially exposed to the shift. The professional networking site, used by more than 1.3 billion registered members to follow companies, industries and workplace news, had a larger share of AI-generated long-form posts than Medium, X or Substack, according to a July analysis by AI-detection startup Pangram.
Pangram classified more than 40% of LinkedIn’s long-form posts as entirely AI-generated.
Key Takeaways
- LinkedIn added a new feature last month: a button users can press if the content they’re seeing “seems like AI slop,” or poorly generated AI content.
- In the first two weeks of the button’s launch, more than a million people have used it.
- AI-detection startup Pangram recently classified more than 40% of LinkedIn’s long-form posts as entirely AI-generated.
LinkedIn doesn’t want to lose users to a flood of AI content.
In an effort to create a better user experience, LinkedIn chief product officer Hari Srinivasan announced in a post on the platform last month that the company would include a new feature: a “seems like AI slop” button.
AI slop refers to poorly generated AI content. LinkedIn is allowing users to flag posts that appear to be AI-generated.