Social networking services are increasingly dividing over how to handle synthetic media, with major platforms taking opposing approaches to the influx of machine-generated text and imagery across user feeds.<\/p>\n
Professional networking platform LinkedIn<\/a> has introduced a direct reporting tool allowing members to flag low-quality content under a new “Seems like AI slop” button in its post options menu. Chief Product Officer Hari Srinivasan explained in a public announcement that user reports will help train feed algorithms to filter out low-effort synthetic posts while assisting users in identifying content that feels inauthentic.<\/p>\n The update reflects growing friction around automated posting on the business-focused network. Analysis by AI detection platform Pangram recently revealed that 40 percent of long-form posts and 30 percent of short-form posts on LinkedIn were entirely generated by artificial intelligence, underscoring the rapid shift in content creation since public access to generative tools expanded rapidly in late 2022.<\/p>\n Alongside the new reporting metric, LinkedIn is discontinuing its automated “enhance your post” feature, replacing it with an AI-powered proofreading tool engineered to edit grammar without altering a writer’s personal tone. The company is also tightening profile and page verification standards to curb coordinated automation. According to Srinivasan, the network blocks hundreds of thousands of automated slop comments every day and has halted billions of large-scale automation attempts in recent months.<\/p>\n Snapchat<\/a> is taking a similar stance against fully automated media. The platform confirmed it will no longer feature fully AI-generated videos inside its Spotlight discovery feed, reserving algorithmically recommended reach for original human creation. While content edited using Snapchat’s native generative tools remains eligible for Spotlight, those posts will feature explicit transparency tags indicating synthetic enhancement.<\/p>\n By contrast, Meta<\/a><\/a> is moving in the opposite direction by building synthetic media generation directly into its primary user experience. The company recently rolled out Muse Image, a model capable of synthesizing images using context clues such as clothing, environments, and specific stylistic traits, alongside a preview of its Muse Video generator.<\/p>\n To mitigate attribution concerns, Meta implemented an invisible watermarking framework titled “Content Seal,” designed to persist on generated media even after cropping or screenshotting, with plans to expand the mechanism to synthetic video. Speaking during Meta’s second-quarter earnings call, Chief Executive Officer Mark Zuckerberg<\/a> stated that automated media generation tools like Muse Image and Muse Video would substantially enlarge the available pool of discoverable platform content to boost user engagement.<\/p>\n