Meta has discontinued Muse Image, an AI feature launched earlier this week that allowed users to generate pictures using any public Instagram account, after the tool drew widespread criticism over privacy concerns from talent agencies, actors and a Hollywood union. The company confirmed the rollback in a statement on Friday, acknowledging that the feature “missed the mark.”
A Feature Built on an Opt-Out, Not Opt-In, Design
Meta had introduced Muse Image on Tuesday as its first image-generation model from Meta Superintelligence Labs, integrated directly into the Meta AI chatbot. The tool let users manipulate images of people simply by @-mentioning any public Instagram account, whether their own or someone else’s, and edit the resulting images through sketches.
The backlash centred on how the feature handled consent. Any user over 18 with a public Instagram account was automatically included in the image generator by default, and had to actively opt out if they did not want their photos referenced this way. Emmy-winning actor Hannah Einbinder, known for “Hacks,” publicly criticised the feature on Instagram after discovering it had been switched on automatically, urging followers to turn it off themselves.
Talent agency CAA, whose clients include Tom Hanks and Meryl Streep, said it raised the issue directly with Meta and pushed for a more reasonable approach. The agency argued that no one’s name, image, likeness, voice or creative work should be used by any third party, including AI models, without clear, documented consent, framing genuine innovation as one that respects creators’ rights rather than handing control to platforms.
Meta’s Reversal and the Pattern It Fits
Meta’s statement described its original intent as offering a useful creative tool that gave people control over whether their public content could be referenced this way, before conceding that the design had failed to achieve that. CAA welcomed the company’s swift removal of the feature following the announcement.
The episode is not an isolated one in the AI industry’s recent history. OpenAI faced a similar controversy over an opt-out likeness feature in its Sora 2 video model earlier this year, drawing comparable criticism before the company changed its policies and eventually shut the feature down altogether. Meta’s decision follows that same trajectory, suggesting AI companies are still calibrating, through repeated public backlash, where the boundary sits between creative tooling and unauthorised use of a person’s likeness.
Why This Matters Beyond the US
The controversy carries direct relevance for India, where the Digital Personal Data Protection Act and its accompanying Rules, notified in November 2025, require platforms to secure clear, informed consent before processing personal data, a standard that an automatic opt-in default for using someone’s public photos would likely struggle to satisfy under Indian law as well. India has also seen its own wave of AI-generated likeness controversies this year, including deepfake videos of public figures used in investment scams, adding to a broader pattern in which AI image and video tools are outpacing the consent frameworks meant to govern them.
For a company already facing scrutiny in India over separate matters, including its proposed WhatsApp username feature and alleged child safety failures on Instagram, the Muse Image reversal adds to a growing list of instances where Meta has had to walk back a product decision after underestimating how users and regulators would react to automatic, low-consent defaults involving personal likeness and data.
