
Meta cancels AI image feature after Hollywood backlash: 'This feature missed the mark'
Meta removed a Muse AI feature on Friday evening that let users generate images by tagging public Instagram profiles, days after CAA and SAG-AFTRA condemned its opt-out consent model.
The feature and its rollout
Meta introduced the Muse Image model on Tuesday, allowing users to manipulate a photo of any person by @-mentioning their public Instagram account. The company framed it as a creative tool built "in service of the social experiences billions of people already love." The feature applied automatically to all public profiles. Users had to navigate settings to opt out. Private accounts and users under 18 were excluded. Once an image was generated, it remained online permanently. The model produced what Meta called "remixes" from the source account's photos.
Hollywood pushes back
Creative Artists Agency issued a statement on Wednesday rejecting the opt-out approach.
No one's name, image, likeness, voice, or creative work should be used by any third party, including AI models, without clear, documented consent.
CAA added that true innovation puts creators first, respecting their rights and giving them real control rather than handing it over to platforms. The guild called for clear disclosures on AI-generated material and swift mechanisms to spot, track, and remove misuse.
SAG-AFTRA followed on Thursday with its own statement and a direct call for members to dig into Instagram's settings and opt out.
Anything other than a clear and conspicuous OPT-IN for these types of uses of Instagram users' images is unacceptable, and an utter miscalculation of public sentiment regarding the obvious dangers and harms inherent in such use.
Meta retreats
By Friday evening the feature was gone. Meta posted an update acknowledging the criticism.
We've heard the feedback that this feature missed the mark, so it's no longer available.
The company said its intent had been to provide a useful creative tool while giving people control over whether their public content could be referenced. It did not announce any replacement or revised timeline.
A pattern of AI backtracks
The retreat bears a striking similarity to the rollout of OpenAI's Sora app in 2025, according to The Hollywood Reporter. Sora launched with limited intellectual property protections and quickly produced infringing characters, including recognizable celebrities. OpenAI pulled back within days and later shut down Sora entirely as it shifted its focus to enterprise customers.
- Meta launches Muse Image model on Instagram, enabling AI image generation via @-mention of public accounts.
- CAA issues statement condemning opt-out consent and demanding clear, documented permission for any use of likeness.
- SAG-AFTRA tells members to opt out and calls the default-in mechanism 'an utter miscalculation of public sentiment.'
- Meta removes the feature, stating it 'missed the mark' and is no longer available.
What the tool actually did
A Muse user needed only to tag a public or unprotected Instagram feed, and the account's images became material for the AI generator to remix. The resulting images were available online permanently. Meta maintained that it built the system with safety guardrails from day one. However, the core objection from both CAA and SAG-AFTRA was architectural: the default was participation, and consent required an active step to withdraw.

