Instagram has begun rolling out a new AI image tool called Muse Image, and the first thing many users will notice is not what it can create, but what it can use: public Instagram profiles. The feature lets people reference a public account when generating or editing an image with Meta’s AI, raising a simple but uncomfortable question for anyone who posts online: should somebody else be able to make an AI image based on you without asking first?

For users who want out, the immediate fix is relatively straightforward. Open Instagram, go to your profile, tap the three-line menu in the top right, then select Sharing and reuse. Find the setting that reads something like “Allow people to reuse your content on Instagram and with AI features at Meta,” and switch off both the Posts and Reels toggles. The exact wording can differ by app version, and the feature is still being rolled out in the US, so some people may not yet see the option.

The bigger issue is that this is an opt-out system. Public accounts are included by default, meaning the burden falls on users to find a setting, understand what it governs, and turn it off after the fact. Meta says the tool is designed as a creative feature and says it has put controls and safety guardrails in place, while private accounts and accounts belonging to under-18s are excluded from the feature automatically.

Still, the backlash is understandable. Instagram is not just a repository for polished photos anymore. For many people, it is a public record of their face, their friends, their workplace, their children, their travel habits and sometimes their location. When a platform makes those images easier to turn into new, synthetic pictures, the line between harmless remixing and unwanted manipulation can get blurry very quickly.

What Muse Image changes

Meta has positioned Muse Image as a new image-generation model available across its ecosystem, including the Meta AI app, WhatsApp and Instagram. It is part of the company’s broader push to put generative AI tools in front of everyday users rather than limiting them to specialist creative software.

The feature’s controversial element is its treatment of public Instagram accounts as a reference point. Someone can tag or name a public profile in an AI prompt and generate an image based on that account’s content, rather than working only from photos they have taken or uploaded themselves. That may sound like a minor technical convenience, but it changes the social contract around public posting.

Public does not necessarily mean freely available for any imaginable use. Plenty of people keep public accounts because they are creators, small-business owners, freelancers, journalists, artists or simply want to share with a wide circle. They may accept that strangers can view and repost their content, but that is different from accepting that their likeness can become raw material for AI-generated images.

There is also a practical limitation worth knowing: turning off the setting is forward-looking. It can prevent future reuse through the feature, but it will not erase images, remixes or AI creations that may already have been made and circulated.

Why consent is central

The consent concern is less about whether every AI image will be malicious and more about who gets to decide. A person could use Muse Image to make a lighthearted birthday image of a friend, but the same capability can be used to place somebody in a situation they were never in, associate them with a message they never endorsed, or produce images that are embarrassing, sexualized or misleading.

This is the same tension that has followed deepfake technology for years. Image generators do not need to produce a perfect impersonation to cause harm. A believable-enough picture can travel through a group chat, a workplace Slack channel or a local Facebook group before the person depicted even realizes it exists.

Meta says it will act on content that breaks its Community Standards, but moderation is inherently reactive. By the time a report is reviewed, copied images may already have been downloaded, reposted or used to target someone elsewhere.

For most users, this is why the opt-out setting matters even if they are not especially concerned about generative AI in general. It is less a statement against the technology than a decision about control. You might be comfortable making your own AI images, but not comfortable with strangers experimenting with your identity.

The scam risk is real

The other concern is fraud. Cybercriminals do not need an AI image tool to run impersonation scams, but tools that make convincing synthetic visuals easier can reduce the time and skill required to create a fake account, a fake emergency, or a fake endorsement.

A scammer could use a public profile as the basis for a convincing impersonation page, then message friends or followers with a familiar-looking photo and an urgent request for money. They could use AI-generated images to make a fake giveaway appear legitimate, fabricate proof of a supposed transaction, or support an impersonation attempt that starts on Instagram and moves to WhatsApp, text or email.

Security firm Malwarebytes warned that giving users an official way to generate AI images based on public profiles lowers the barrier for impersonation, scams and other forms of abuse. The important phrase is “lowers the barrier.” The risk is not that every public Instagram user will suddenly become a fraud victim. It is that the cost of creating convincing visual deception keeps falling.

That makes old-fashioned verification more important, not less. If a friend messages asking for money, gift cards, crypto, a login code or an urgent favor, do not rely on the profile photo or an apparently personal image as proof. Call them, message them through a different channel, or ask a question only the real person would know. Treat urgency as a warning sign, particularly when money or account access is involved.

A sensible privacy reset

Switching off the Muse Image reuse setting is a good first step, but it is not the only one worth taking. If you want the strongest protection available within Instagram’s current approach, making your account private prevents strangers from using your public profile as source material for the feature. That is a more consequential choice for creators and businesses, of course, because it limits discoverability too.

For public-facing users who cannot or do not want to go private, it is worth auditing what is already visible. Review old posts and Highlights. Remove location details from content that does not need them. Be selective about images showing children, work badges, travel itineraries or information that could help someone build a convincing impersonation story.

It is also a good moment to enable two-factor authentication in Meta’s Accounts Center (now just “Meta Account”). This will not stop somebody from creating an AI image inspired by your public profile, but it can make it far harder for an attacker to take over your actual Instagram account and use it to make a scam more credible. An authenticator app is generally preferable to SMS-based codes, though either is stronger than relying on a password alone.

Muse Image is not the first tool to blur the boundary between sharing an image and surrendering control over how it might be repurposed. But its arrival on Instagram brings that question to a platform where faces, identities and social connections are already tightly intertwined. Turning off the setting will not solve the deepfake problem. It does, however, give users one clear way to say that a public post is not an open invitation for anyone to generate a new version of them.