Meta turns off the Instagram feature that let users make AI deepfakes of public accounts
TL;DR
Meta is shutting down the new Instagram feature that let users generate AI images based on public accounts simply by tagging them with an @-mention inside Meta AI. The original setup allowed content from any public Instagram account to be used in AI generations without explicit permission from the account owner. An opt-out existed, but it was buried in settings.
Nauti's Take
This was not a small UX mistake but a basic product failure: public does not mean fair game for AI remixing. Shipping a feature like this while hiding protection behind an opt-out menu effectively bakes abuse into the rollout.
Meta moved only once the backlash became loud enough, which means the shutdown alone is not evidence of trustworthiness. The real test is whether future AI products launch with explicit consent and hard guardrails instead of retroactive damage control.
Briefingshow
This case shows how fast a GenAI feature collapses when a platform treats public visibility as implied consent. For creators, celebrities, and ordinary users, the issue is not just copyright but identity protection, abuse, and reputational harm. The fact that Meta only backed down after public pressure is a sign that internal risk checks around these launches are still too weak.
Sources