Instagram Photos Hijacked—Opt-Out Won’t Save Past

Meta has quietly enrolled every adult with a public Instagram account into an AI image tool — without telling them — and photos already generated using their likeness cannot be deleted even after opting out.

Story Snapshot

Meta’s new Muse Image tool automatically uses public Instagram photos to generate AI images — no user consent or notification required.
The opt-out setting is buried deep in Instagram’s app settings, and most users have no idea it exists.
Even after opting out, AI images already made with your photos are not removed.
Critics warn the tool makes it easier to create fake or non-consensual images of real people.

What Meta’s Muse Image Tool Actually Does

Meta launched a new AI image tool called Muse Image in July 2026. It lets anyone tag a public Instagram account and generate AI images based on that person’s photos. Every adult with a public Instagram account was automatically enrolled — no warning, no permission asked. Meta did not notify users when their photos were used. The tool is currently live for U.S. users only.

The tool works through Meta’s AI app. A user types a prompt and tags a public Instagram handle. The system then pulls from that account’s public photos to build the image. Meta says it includes an invisible watermark on each generated image, and a detection tool is available at its website. But critics say that does little to stop misuse once an image is out in the wild.

The Opt-Out Is Hidden — and It Doesn’t Undo the Damage

Meta does offer a way to stop your photos from being used going forward. But finding it takes effort. Users must open Instagram, go to settings, find the “Sharing and reuse” tab, and turn off a toggle for Posts and Reels. Wired described it as having to “dig into the app’s settings.” Most people will never find it — and Meta never told them to look. That’s a major problem.

Even worse, opting out does not fix what’s already done. Instagram’s own Help Center states clearly: “already existing AI images made with your content will not be deleted.” So if someone already created a fake image of you, it stays out there. You have no way to remove it. That is not real user control — it’s a fig leaf.

Privacy Advocates Sound the Alarm

Donald Campbell, advocacy director at the nonprofit group Foxglove, called the opt-out model an “obvious recipe for disaster.” His concern is straightforward. When a platform enrolls millions of people by default and buries the exit door, most users never escape. Critics also point out the tool makes it easier to create non-consensual altered images — fake photos that put real people in situations they were never in.

Meta argues the feature fits within Instagram’s existing content reuse policy, which already allows certain uses of public posts. The company says private accounts and users under 18 are excluded. A spokesperson called the opt-out “just a couple clicks.” But that framing ignores the core issue: users were never asked. They were simply enrolled. Telling someone they can leave after you’ve already taken their data is not the same as asking permission first. This is exactly the kind of quiet corporate overreach that erodes trust — and it deserves far more scrutiny than it’s getting.