Meta has disabled its Muse Image tagging feature less than one week after launch, following public backlash over a mechanism that allowed any user to generate AI images referencing any public Instagram account simply by @-mentioning it, with no consent required from the account owner.
The feature was part of Meta's newly announced Muse Image AI model. As designed, it gave third parties the ability to feed public creators, brands, and individuals into AI image generation without opt-in controls. Meta framed the rollback as a gap between intent and execution, stating the goal was a 'useful creative tool,' but the structure of the system made non-consensual deepfake-style outputs trivially easy to produce at scale.
The full Verge piece is worth reading for the specifics of how the feature was architected, what Meta's blog post update actually says versus what it omits, and what this episode signals about how the company is stress-testing AI product boundaries in public before locking down policy.
[READ ORIGINAL →]