Meta shut down the tagging function inside Muse Image, its new AI image tool, within days of launching it on Instagram. The company said it had “missed the mark” and the feature was “no longer available,” according to BBC News reporter Kali Hays.
Muse Image let users of the Meta AI chatbot tag any public-facing Instagram account and pull that account’s photos into an AI generation or editing session. Because Instagram users were enrolled by default, anyone with a public profile could have their likeness turned into raw material for someone else’s AI image without being asked first.
That default-on design is the entire story. Muse Image did not malfunction. It executed precisely what Meta built: a system that treats a public account as a queryable dataset rather than a person’s property. The backlash was not a bug report. It was a rejection of the premise that visibility on a platform equals consent to be remixed by anyone who opens a chatbot.
SAG-AFTRA, the Hollywood union representing actors and performers, called Meta’s reversal a “win.” The union had told members and “all Instagram users” to act to protect their likeness, describing the launch as an “utter miscalculation of public sentiment regarding the obvious dangers and harms inherent in such use,” per BBC News.
Privacy International, the London-based human rights charity, went further, telling the BBC the episode was “the latest sign AI companies see people’s images and data as raw material to be exploited.” That framing does not treat Muse Image as an isolated misstep. It treats the feature as evidence of a pattern: opt-out by default, correction only after public pressure.
Meta’s own explanation supports that reading. “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,” the company said. “We’ve heard the feedback.” Read plainly, that is an admission that the control mechanism shipped broken, or did not ship at all, and that Meta’s internal review treated an opt-in requirement as optional rather than foreseeable.
Pulling a live feature days after a broad rollout is expensive. It means engineering, legal, and policy teams signed off on default consent for likeness use, and only public objection caught what internal review missed. That gap, not the model’s technical performance, is the failure worth tracking.
Muse Image was Meta’s first foray into AI image generation, and the company has said more AI features are planned for WhatsApp, Facebook, and Messenger, alongside a video tool already in development. Meta declined to comment further to the BBC.
Any product team building AI editing on top of user-generated content should treat this as the template to avoid: default participation without explicit, revocable consent draws the same reaction regardless of how capable the underlying model is. The binding constraint on this product category is not generation quality. It is whether the people whose images feed the system ever agreed to that use.
According to BBC News, in reporting by technology reporter Kali Hays published July 11, 2026.
