Meta Pulls Instagram AI Feature That Generated Images Using Others' Likeness

Meta pulled Muse Image from Instagram on Friday, an AI image generator that let users create pictures using the likeness of any public account, simply by tagging it with an @ mention. The decision came just days after the tool's launch, as a wave of criticism swept in from both ordinary users and agencies representing celebrities.

Muse Image was pitched as a creative tool for generating images inspired by public Instagram accounts. In practice, that meant any user could generate a picture referencing another public person's appearance, without that person's knowledge and without any notification that their likeness had been used.

Opt-out by default, hard to turn off

At the heart of the controversy was the feature's default behavior. Rather than asking users for consent to use their photos, Meta built an opt-out model: every account with a public profile was automatically available to the generator until its owner found the setting to disable that option. For many creators and everyday users, that meant learning the feature existed from social media chatter rather than from any notification sent by Instagram.

Critics warned that this kind of design opens the door to abuses resembling already-known cases of AI being used to create unauthorized, altered depictions of public figures. CAA, which represents some of Hollywood's biggest names, directly pressed Meta to change its approach.

Meta's quick reversal

The company backed down faster than in many previous disputes over privacy and AI. In a statement published Friday, Meta acknowledged the feature fell short of expectations and announced it would be removed from the platform entirely, without giving a date for any potential return in a revised form.

Our intent was to provide a useful creative tool and give people control over whether their public content could be used this way. We've heard feedback that this feature missed the mark, so it's no longer available - Meta statement

Another chapter in the AI likeness dispute

This marks the second time in recent weeks that Meta has had to answer for AI-generated content built on Instagram users' likenesses without clear consent. The episode shows that regulatory and public pressure around the use of other people's likeness in image generators is mounting faster than tech companies can anticipate in pre-launch testing.

For Polish Instagram users, the incident is a reminder that default privacy settings on public accounts can get swept up in new AI features before people even learn about them through official channels. The case could also accelerate discussion within the European Union over whether image generators built on other people's social media profiles require explicit upfront consent, rather than an after-the-fact opt-out.

Sources: TechCrunch (techcrunch.com), Variety (variety.com), Fox 10 Phoenix (fox10phoenix.com)