Powered by Smartsupp
Site navigation
Store catalog
Digital goods marketplace

No Search or Recommendations: How Meta’s Algorithms Penalize Undisclosed AI Use

Meta has updated the rules for pages that use virtual characters. Instead of the previous «AI creator» badge, the social network introduced an «AI-generated profile» label. Authors are now required to manually enable a toggle in the settings indicating that the profile was created using AI, otherwise their posts will stop receiving reach.

Why the rules were tightened

The trigger was media investigations. Wired journalists uncovered a network of generated models that chatted with users in DMs and redirected them to the Goose dating service. At the same time, The New York Times published a story about fake doctors who gave medical advice in comments and sold dietary supplements.

Following these reports, Meta abandoned the old «AI creator» label, which allowed generated characters to be presented as real creators using video filters.

What happens without the label

The platform’s algorithms have learned to detect signs of generative AI in photos and videos. If the author does not apply the system label, the profile is subject to restrictions:

  • Videos stop appearing in Reels recommendations and the Explore tab;

  • Posts stop being shown in the feed even to existing followers;

  • The page is removed from internal search and the list of similar profiles;

  • After the first user complaints, the page is sent for a ban.

Meta is closing off the ability to run generated profiles while presenting them as real people. Authors will now have to either openly use the label and lose the trust of some users, or completely abandon virtual characters in favor of filming with real people.