We Heard the Feedback

Meta pulled the @-mention feature that let strangers composite any public Instagram face into Muse Image — three days after launch — after Hollywood unions and privacy advocates forced a retreat the company framed as listening to users.

What's new: On July 10, Meta discontinued the Instagram tagging tool inside Muse Image — not the whole model. Users could type @username and pull a public account's photos into AI generations without that person's knowledge or consent.

• Meta launched Muse Image on July 7 through Meta Superintelligence Labs, per Reuters.
• By July 10, Meta admitted in a statement: "We've heard the feedback that this feature missed the mark, so it's no longer available."
• The core Muse Image generator — text-to-image, sketch edits, WhatsApp and Stories integration — remains live, per the BBC.
• Meta declined further comment to the BBC after the reversal.

Why it matters: Unions and creators forced one of the fastest Big Tech rollbacks on likeness consent in years — not a buried settings change.

• SAG-AFTRA urged members to opt out on July 9, calling anything less than "a clear and conspicuous opt-in" "unacceptable" and "an utter miscalculation of public sentiment regarding the obvious dangers and harms inherent in such use," per Reuters.
• On July 10, the union welcomed the pull: "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."
• CAA demanded Wednesday 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," per Deadline — then Friday commended Meta for putting "individual rights and consent at the forefront."
• On Instagram, SAG-AFTRA posted: "A win is a win."

Who pushed back: It wasn't just guild lawyers.

• Emmy winner Hannah Einbinder (Hacks) criticized the feature on Instagram, saying it had been turned on automatically and urging users to disable it, per Reuters.
• Privacy International told the BBC the episode was "the latest sign AI companies see people's images and data as raw material to be exploited."
• WIRED reported public profiles were automatically opted in with no notification when strangers tagged them — and updated its story July 10 to note Meta rolled the tagging feature back three days after release.

What Meta kept: The retreat is narrower than the headlines suggest.

• Muse Image still generates images from text, blends photos you upload yourself, and edits via sketch — free in Meta AI, Instagram Stories, and WhatsApp.
• Muse Video remains in preview on the same Superintelligence Labs stack.
• Advantage+ advertiser access is still "in the coming weeks," per Meta's July 7 newsroom post.
• Facebook and Messenger integrations are still planned.

The pattern: Meta ships consent as opt-out, waits for backlash, then quotes the backlash back at users.

• Meta's own policy said people "will not be notified" when their public content is used in someone else's AI image — control existed, silence was guaranteed.
• The company resumed training on EU users' public posts in April 2025 only after Ireland's Data Protection Commission paused the rollout for a year.
• UK regulator Ofcom is already investigating X over Grok's role in non-consensual synthetic images — the same harm vector unions warned Muse would multiply.

What we cannot verify:

• How many @-mention generations ran during the three-day window before the kill switch.
• Whether Meta will relaunch the feature under opt-in framing — the company gave no return timeline.
• Whether images created during the window remain downloadable or shareable off-platform.

Convina's view: Meta didn't kill Muse Image — it killed the part users could actually see fighting over. That is progress, and SAG-AFTRA earned it. But a three-day rollback after a default-opt-in launch is not a policy change; it is damage control. The model, the video preview, and the ad pipeline are still coming. Until regulators mandate opt-in consent and real-time notice — not buried Instagram toggles — the next @-mention will just ship under a different name.