LinkedIn and X Are Flooded with AI Spam, Browsing Data Suggests (www.404media.co)

A new analysis by Pangram, a company specializing in detecting AI-generated content, reveals that an alarming proportion of posts on social media platforms like LinkedIn and X are AI-generated. The study found that approximately 41% of longform content on LinkedIn and about a third of lengthy posts on X are fully AI-written, while around 10% of similar content on Reddit and Substack falls into the same category. This data, derived from nearly a million posts viewed by users of Pangram's Chrome extension, is significant as it underscores the pervasive nature of AI-generated content in professional and social discourse, raising questions about the integrity of online communications. The implications are profound for the AI/ML community and platform moderators. As AI-generated content becomes more common, especially in longform contexts, it challenges the authenticity of user interactions and may lead to a diminished reading experience. Platforms like LinkedIn and Reddit are attempting to curb this trend; LinkedIn has removed AI writing tools from its post features, while Reddit is actively banning AI-generated spam to prioritize human engagement. With users increasingly encountering AI-generated "slop," the implications for content quality and platform trustworthiness are becoming critical issues that need addressing in the rapidly evolving landscape of social media and AI.