Meta Pulls Back Image Feature From Instagram, Here’s Why!

California: Meta has shut down a controversial AI image feature on Instagram just three days after launching it, following a fast-moving backlash from talent agencies, unions, and privacy advocates over how it handled consent.

The feature was part of Muse Image, Meta Superintelligence Labs’ first publicly released image generation model, positioned as the second major output from the division led by Alexandr Wang, who previously founded Scale AI, following an earlier large language model release in April. Muse Image let users edit photos through sketches or annotations, remove people from the background of images, render legible text inside pictures, and generate scannable QR codes, all through conversational prompts.

Unlike typical text-to-image generators that run a single forward pass, Muse Image reasons before, during, and after generating an image, calling on tools like web search and code execution and refining its own drafts along the way.

The feature that triggered the uproar allowed users to bring another person into a generated image simply by tagging their public Instagram handle. Every adult with a public account was included by default, and users had to manually dig through settings to block others from remixing their photos, while private accounts and users under 18 were automatically excluded. Critics quickly pointed out that the mechanism was structurally set up to be misused.

The design made the backlash almost inevitable. The opt-out toggle sat several menu levels deep inside settings, so users had to already know it existed to find it, and disabling it only stopped future image generation, it did nothing to remove images that had already been created before someone opted out. According to a privacy researcher cited by Tech Times, some users reported the toggle wasn’t even visible in their version of the app when the feature first went live, meaning parts of the rollout were active before the opt-out mechanism was fully in place.

Opposition mounted fast. Creative Artists Agency, whose client roster includes major names in Hollywood, argued that no one’s identity or creative work should be used by outside parties, AI systems included, without clear and documented permission. SAG-AFTRA, representing roughly 160,000 media professionals, urged its members to actively protect their likeness by adjusting their settings.

Privacy International told the BBC the episode reflected a broader pattern of AI companies treating people’s images as raw material. Even India’s Ministry of Electronics and Information Technology reportedly weighed in demanding a reversal.

Meta reversed course on Friday. In an update to its original announcement, the company acknowledged that the feature “missed the mark” and said it was no longer available. It framed its original goal as giving people a useful creative tool while letting them control whether their public content could be referenced this way.