Meta built a powerful artificial intelligence tool that let any stranger take your public Instagram photos and turn them into fake images without your permission. Users were dropping unsuspecting people into fake adverts, altered photos and bizarre situations they never agreed to, simply by typing a description and tagging their account. Within just three days, the massive backlash from furious users and Hollywood stars forced the tech giant into one of the fastest U-turns in its history.
The feature, called Muse Image, launched on Tuesday as Meta's big new creative tool. By Friday afternoon it was completely shut down, with Meta admitting in a brief statement that the feature 'missed the mark, so it's no longer available'.
The Secret Settings Trap
The way the tool worked was quite simple. Anyone could type a phrase like 'astronaut on Mars' followed by your Instagram username, and the AI would instantly scrape your pictures to build your face into the scene. Every public account belonging to an adult was automatically opted in. Turning it off required digging deep into the app's settings menu to find a hidden toggle switch, and even then, any fake images already created by strangers remained online.
Worse still, Instagram never sent a notification when someone used your face. Many everyday creators only discovered they had been cloned when their own followers spotted the fake images on strangers' profiles and sent alerts to warn them.
No Protection for Normal Users
Hollywood power players quickly stepped in to fight back. Top talent agencies representing stars like Tom Cruise and Charlize Theron warned that nobody's face or voice should ever be used by AI without clear, signed permission.
Actors' unions called the move a massive mistake that ignored the obvious dangers of identity theft and online abuse. The backlash highlighted a scary reality, because if rich celebrities with massive legal teams were left vulnerable, ordinary people posting holiday photos stood absolutely no chance.
Meta tried to defend the tool by saying that private accounts were safe and that the AI images contained a hidden watermark. Neither excuse fixed the main issue. A hidden watermark only labels a fake after it has already spread, and doing this did nothing to protect the hundreds of millions of ordinary adults who never wanted their faces used in the first place.
A Lesson Learned the Hard Way
Meta should have seen this disaster coming. The tech world is already facing massive scrutiny over these exact dangers, especially after recent high-profile scandals where rival AI tools were used to create abusive and sexualised fake images of real people. Meta had these warnings right in front of them and chose to push ahead anyway.
The company now says it will bring back a limited version of the tool without the ability to tag other people. The sudden retreat proves that the public is no longer willing to be a lab rat for big tech. Silicon Valley can no longer treat your personal photos as free property, because users have finally found their voice and will force a retreat within days.
