Meta kills Instagram AI image feature after four-day backlash over privacy and consent
The Muse Image tool, which let users generate AI deepfakes from any public Instagram account without permission, was disabled late Friday after criticism from SAG-AFTRA, CAA, and thousands of users.
What the feature did
Meta unveiled Muse Image on Tuesday as its first image-generation model from Meta Superintelligence Labs. Integrated into the Meta AI chatbot, the tool allowed any user to @-mention a public Instagram account and instantly generate AI images based on that account's posts. Users could also edit generated images through sketches. The feature was part of a larger suite of AI tools announced this week, including special filters created by Muse Image that remain available on Instagram. The image-from-account capability was switched on by default for every public profile, with no notification sent to account owners when their content was used.
The consent problem
Users could only prevent their likeness from being fed into the generator by digging into the Settings menu and toggling off an option labelled "Allow people to create with and reuse your content," or by setting their entire profile to private. Haley McNamara, executive director of the National Center on Sexual Exploitation, described the high-risk design as unacceptable, noting it put the onus on individuals to jump through hoops to opt out. The feature was immediately flagged as a potential tool for sextortion and scammers.
Not only does this obviously erode our rights to our own likeness, but it is an obvious tool for #sextortion and other scammers!
Hollywood intervenes
The actors' union SAG-AFTRA urged members on Thursday to manually opt out and condemned the automatic enrolment as an utter miscalculation of public sentiment. Creative Artists Agency, whose clients include Tom Hanks and Meryl Streep, contacted Meta directly. Emmy-winning actor Hannah Einbinder criticised the feature on Instagram, telling her followers it had been turned on automatically and instructing them to switch it off. CAA issued a statement insisting that no person's name, image, likeness, voice or creative work should be used by any third party, including AI models, without clear, documented consent.
With the dangers of nonconsensual digital replicas well known to all, a feature that encouraged that behavior is unwise. We appreciate its discontinuance. It is the responsible thing to do.
Meta's reversal
Late Friday, Meta updated its blog post and posted on Instagram to announce the feature was no longer available. The company said its intent was to provide a useful creative tool and to give people control over whether their public content could be referenced, but that feedback showed the feature missed the mark. Puck News founding partner Dylan Byers first reported the decision. Meta confirmed the Muse Image model itself remains active on WhatsApp and the standalone Meta AI app, while the @-mention image-generation capability on Instagram has been entirely disabled.
- Meta launches Muse Image on Instagram with automatic opt-in for all public accounts.
- Users and outlets publish guides on how to disable the feature; backlash begins.
- CAA issues statement demanding clear consent for use of likeness and contacts Meta directly.
- SAG-AFTRA urges all members and Instagram users to opt out; Hannah Einbinder speaks out.
- Meta announces the feature is no longer available, updating its blog post.
Broader AI backlash
The incident is the latest in a series of AI rollouts that have drawn sharp criticism over copyright and privacy. OpenAI faced similar concerns when it released its Sora video generator in September 2025 and eventually shut the app in March 2026. Elon Musk's X blocked its Grok chatbot from posting certain images after millions of manipulated images of real women and children appeared on the platform this year. Google and other companies have also faced pushback over how their AI systems create images. The New York Times, which reported on Meta's decision, is currently suing OpenAI for copyright infringement related to AI training.
