Meta enabled a feature allowing users to generate AI images using photos from public Instagram profiles without explicit consent from the people pictured. Privacy advocates immediately condemned the move as dangerous.
The company claims users can opt out through their settings, but privacy campaigners argue the opt-out mechanism shifts responsibility unfairly onto individuals rather than requiring opt-in consent upfront. Privacy International and other groups called the approach a "recipe for disaster," warning it creates pathways for deepfakes, harassment, and non-consensual imagery.
Meta's AI image generation tools, powered by its Llama technology, now tap into public profile photos as training material and source images. The feature essentially democratizes access to someone's likeness without their knowledge or permission, then burdens them with finding and disabling the setting after the fact.
The backlash reflects broader anxiety around AI's use of personal data. Instagram users, particularly women and minors, face heightened risks of non-consensual synthetic media. A user's public profile picture, posted for networking or social purposes, now becomes raw material for generating potentially compromising or defamatory AI images.
Meta's defense rests on the opt-out option and the fact that the profiles are already public. But privacy advocates note this reasoning ignores the difference between consenting to a photo's visibility and consenting to its use for AI generation. Public does not mean free to synthesize.
The controversy lands Meta amid ongoing regulatory scrutiny in the EU and elsewhere over AI practices and data usage. This move may accelerate pressure for stronger AI consent laws requiring explicit opt-in before using personal biometric data for generative purposes. For Instagram users, it signals another erosion of control over their own image in the age of synthetic media.