Meta just pulled an AI image-making feature from Instagram less than a week after it launched. The company called it a mistake, saying the tool "missed the mark, so it's no longer available." BBC

What the tool did

The feature was called Muse Image, and it worked like this: you could tag someone else's public Instagram account, and the AI would use their photos to create or change images. Meta said the point was to let friends collaborate on creative projects — like reimagining photos together. Meta AI

Muse Image is Meta's most advanced image-generation model to date. It can follow detailed instructions, make precise edits, and combine elements from multiple photos. The company also released a companion video tool at the same time, and both tools build on work Meta started developing years ago. Meta AI

Why people were upset

The controversial part was the tagging feature — the ability to pull someone else's photos without their explicit agreement. Instagram already has other AI tools that let you edit your own photos or change how they look. But this new tool was different: it let one person use another person's account as raw material, even if that second person didn't directly agree.

Meta's own statements suggest the company knew this could be a problem. They talked about giving people "control over whether their public content could be referenced" — but that control didn't seem to work very well in practice. Meta

The bigger picture

Here's what makes this moment worth paying attention to: Instagram is built on people sharing photos and videos publicly. That makes it perfect material for training AI systems. But that same public-by-default setup means when one person wants to use another person's image, it raises a real question: just because something is public doesn't mean everyone is okay with it being used this way.

This isn't really about one feature disappearing. Meta has plans to add more AI tools to WhatsApp, Facebook, and Messenger, with video generation coming down the line. BBC The questions about consent and how people's public presence can be used — those are going to come up again. Pulling this one tool doesn't solve the underlying problem; it just postpones it.

Meta hasn't said whether it will build a new version of the tagging feature, or explained exactly what complaints led to the shutdown. This is part of a larger pattern in the AI industry: companies launch features quickly, real users find problems the designers didn't anticipate, and then companies pull back or change course. Whether that's the right move depends partly on whether companies apply strict scrutiny before launch, or only after the public pushes back.