Meta Pulls Muse Image Feature After Privacy Backlash
Meta removed a Muse Image feature that let people use public Instagram accounts as visual references. The image model itself remains available.

Meta has removed a new AI feature that let people use public Instagram accounts as references for generated images. The company pulled the feature within days of launching Muse Image after criticism about consent and the use of people's likenesses.
Muse Image itself is still available. The change affects one feature inside it, not the full image model.
What Meta removed
At launch, Meta's Muse Image launch announcement said people could tag an Instagram username and let Meta AI use that account's public photos to make a new image. Adults with public accounts could turn the feature off in their settings.
That setup meant the default mattered. A public account could be used as a visual reference until its owner opted out.
The Associated Press reported that Meta removed the feature on Friday. Meta told the AP that it wanted to give people control, but that the feature had missed the mark after the company heard the feedback.
The concern was not just about famous people. Anyone with a public adult account could have been referenced. Actors and creators had an added reason to care because their face and style may be part of how they earn money.
SAG-AFTRA, the union that represents actors and other performers, welcomed the decision. The union said tools should not encourage the use of a person's likeness without consent.
What did not change
Meta did not withdraw Muse Image. The broader model can still create and edit images in Meta AI. Meta's Muse Image technical overview says it is available in the Meta AI app and on meta.ai, with features also reaching Instagram and WhatsApp.
That distinction matters. Meta removed the shortcut that connected a public Instagram account to a new AI image. It did not say that it had stopped the model, its other editing tools or its wider rollout.
This was also a reference feature used while making an image. The available sources do not establish that Meta used these public-account photos to train Muse Image. Calling it a training-data reversal would go beyond the evidence.
Why the fast reversal matters
Meta built Muse Image around its social advantage. Rivals can generate pictures, but Meta owns large networks of people, photos, creators and advertisers. Connecting those networks could make its AI tools feel more personal.
The reversal shows the limit of that advantage. Public content is easy for other people to see. That does not mean account owners expect a product to turn their photos into a one-step input for synthetic images.
For investors tracking who owns Meta or the live META research page, the immediate effect is narrow. One feature disappeared, while the main image model remains. The bigger question is whether Meta will bring back social-reference tools with opt-in consent, stronger notices or tighter limits.
The next useful evidence will be a new product setting, a revised consent design or an updated launch page. Arkolith will update this story if Meta changes the feature again.
Arkolith provides source-linked public information for educational and informational use. This article is not investment advice.
More from Arkolith
Micron Raises Its U.S. Chip Bet to $250 Billion
Micron says it now plans to invest more than $250 billion in U.S. manufacturing and research through 2035 as its first New York fab moves into vertical construction.
Mira Murati’s AI Lab Bets on Models Users Can Train
Thinking Machines Lab says its strategy is to let people train model weights, shape AI with private knowledge and work through live multimodal interfaces.
Apple Sues OpenAI Over Alleged Hardware Secrets
Apple alleges that OpenAI and two former Apple employees misused confidential hardware information. OpenAI says it has no interest in other companies' trade secrets.