SAN FRANCISCO – Meta pulled its Muse AI feature from Instagram less than a week after launch on Thursday, after talent agencies, photographers, and public figures raised immediate objections to a product that let users generate AI images featuring real people’s public Instagram profiles without notifying those people. The feature went live during the week of July 7 and was gone by July 10.
The company’s statement acknowledged the reversal without fully accounting for why the feature passed internal review. “Our intent was to provide a useful creative tool and to give people control over whether their public content could be referenced in this way,” Meta said. “We’ve heard the feedback that this feature missed the mark, so it’s no longer available.”
That framing, positioning the problem as missed execution rather than concept, does not quite fit the available facts. The consent question is not a detail that surfaced only during testing. Using real people’s likenesses without their knowledge to generate synthetic images is a problem that courts, regulators, and the entertainment industry have explicitly grappled with for at least three years. That Meta built and shipped Muse without resolving it suggests either that internal review processes failed to escalate the objection, or that they did and the launch proceeded anyway.
CAA and other major talent agencies scrutinized the feature within days of its appearance, according to TechCrunch, which reported on the backlash and published a guide explaining how users could opt out on July 9. For represented talent, unauthorized AI-generated likenesses present both a rights violation and a potential contract complication, since many talent agreements include image and likeness clauses that agencies administer. An AI tool that bypasses those structures is not, from an agency’s perspective, a product-design problem. It is a legal and commercial threat.
For individual creators and public figures outside the agency ecosystem, the issue is simpler and harder to remediate. When someone with a public Instagram account discovers that strangers have used their face and identity in AI image prompts they never consented to, the standard opt-out mechanism arrives after the fact. The images generated in that window do not disappear because Meta removed the product.
Meta Superintelligence Labs, the company’s AI research division, developed Muse as part of a broader push to integrate generative AI into Instagram’s creative toolset. The product followed Meta’s AI-generated stickers, background removal, and comment-response features, all positioned as quality-of-life additions for casual users. Muse was different in that it introduced a third party, any public account, without that party’s participation. Eastern Herald covered the initial questions the feature raised when it first appeared, in reporting on how Meta Muse used Instagram images in AI generation without consent.
Meta has not said whether it plans to reintroduce a modified version of Muse with explicit consent mechanisms, or whether the removal is permanent. Given that Meta has invested substantially in Meta Superintelligence Labs and publicly framed AI features as central to Instagram’s competitive position against TikTok and Snapchat, a permanent discontinuation would represent a significant course correction.
The episode adds to a pattern in which large AI platforms ship consumer-facing features and then retreat when public or industry pressure reaches a threshold. The broader AI development landscape has grown increasingly contested: with Apple suing OpenAI over alleged trade secret theft, the field’s competitive and legal stakes have rarely been higher. The difference in the Muse case is that the harm being contested was not rival companies poaching technology. It was real people’s images and identities being used without their knowledge.
Federal legislation covering AI likeness rights remains stalled in Congress, where competing industry interests and disagreements over scope have slowed action. In the absence of federal standards, states including California and New York have advanced their own measures, and the entertainment industry embedded consent requirements directly into union contracts following the 2023 strikes. Meta’s Muse, in that context, was not operating in a legal vacuum. The rules, though not yet fully statutory, were visible to anyone paying attention.
The broader question the episode raises is whether the pattern, build a feature, ship it, respond to backlash, remove it, will eventually produce faster scrutiny at the product design stage. Or whether large platforms will continue to treat post-launch public pressure as the effective regulatory mechanism in the absence of binding federal rules. Meta’s statement does not resolve that question. It simply marks the latest instance where the answer arrived after the product was already in users’ hands.
Meta had not responded to requests for additional comment on the timeline of the internal review process by the time this article was published.
