Meta has built a way to turn public Instagram accounts into raw material for artificial intelligence, and it did not ask the people in those accounts first. On July 7, 2026, the company launched Muse Image, described in Meta’s own announcement as the first image generation model from Meta Superintelligence Labs. It is a feature inside the Meta AI app that lets anyone type a prompt, @ mention a public Instagram username, and receive freshly generated pictures of that person. The photos Meta draws on are the ones already sitting on the mentioned account. The person being rendered does not have to approve the request, and given how the tool is built, they are not even told it happened.
Meta pitches the tool as a leap in creative power. The company says Muse Image uses reasoning to interpret complex prompts and can blend multiple photos into a single image that users can download and share to a chat, story or feed, and it ships with more than 30 AI powered effects for Instagram Stories. Meta has positioned it as free for everyday creation in the Meta AI app, with more generation available through the company’s subscription plans. The model already powers creative experiences on Instagram and WhatsApp, is coming to Facebook and Messenger, and will reach advertisers through Meta Advantage+ creative. That reach is part of the point Meta is making, and it is also why the consent question lands as hard as it does.
The consent model is the part drawing the sharpest reaction. Muse Image works on an opt out basis rather than an opt in one. If an Instagram account is set to public, it is included by default, with no consent request sent beforehand and no notification sent afterward. Meta’s own help page states plainly that “You will not be notified about content created using AI features at Meta.” In practice that means a stranger can generate an image of someone they have never met, and that someone may never learn the picture exists.
Reporters who tested the feature found precisely that. Testing Muse Image, they generated images of people they had never followed or interacted with, pulling from public accounts without any of those people granting permission, and the launch drew immediate pushback from users, as TechCrunch reported the day the tool went live. The friction that most users would expect, some prompt asking whether a person is comfortable being reimagined by a stranger’s text box, simply is not there. The account being public is treated as the answer to a question that was never actually asked.
There is a way to step out of the pool, though it takes some hunting to find. To remove an account, a user opens Instagram and taps through to their profile, then opens the menu shown as three lines and goes to the section called Sharing and reuse. There, under a setting labeled “Allow people to reuse your content on Instagram and with AI features at Meta,” the person can toggle off both Posts and Reels. Turning off both is what pulls that account’s photos out of what Muse Image is allowed to reach.
The opt out carries an important limit that is easy to miss. Switching those toggles off only blocks future generations. Anything Muse Image has already produced from an account stays in existence, sitting wherever it was created and saved. A person who discovers the feature after the fact, opts out, and assumes the matter is closed would be wrong on the second count. The setting stops the flow going forward. It does not reach back and undo what has already been made.
Critics argue the whole arrangement inverts how consent is supposed to work. Foxglove, a tech justice nonprofit, called Muse Image an “obvious recipe for disaster,” pointing to a year of documented harms caused by non consensual AI altered images. Their objection is not abstract. When the raw material is real people’s faces and the default is inclusion, the tools that generate flattering portraits are the same tools that can generate images the subject would never have agreed to, and the person in the picture has no say at the moment it is made. The criticism now reaches into Hollywood as well. Creative Artists Agency, the talent firm known as CAA, publicly called out Meta and Instagram for building the Muse tool as opt out rather than opt in, a pointed objection from a firm that represents the actors, athletes and creators whose likeness is exactly what a tool like this can reach.
Meta’s framing runs the other way. The company points out that Muse Image only draws on content people have already chosen to make public, the same photos any Instagram user could scroll through today, and that anyone who objects has an available opt out they can reach in a few taps. Meta says users “have control,” pointing to the settings that let a person disable this use of their pictures, and Mark Zuckerberg publicly pushed back on the privacy concerns, according to reporting from Deadline. From that angle, the feature is doing openly and at scale what was already technically possible with a public profile, and it hands users a switch to turn it off. Those points are worth stating fairly. They also do not change the two facts at the center of the story. The default is opt out, and the people whose photos feed the tool are not notified when their likeness is used.
That gap between what is technically public and what people expect to happen to their photos is where this fight will sit. Most users understand a public account to mean their pictures can be seen. Fewer would assume it means a stranger can feed those pictures into a generator and produce new images of them, quietly, with the burden falling on them to go find a buried toggle and switch it off. Meta has decided that setting the default to inclusion is acceptable as long as the exit exists. Whether regulators, and the people showing up in images they never approved, agree is the question the launch has now forced into the open.
Your First 10 AI Skills
10 practical AI skills, copy-paste prompts and a 7-day plan to start using AI with confidence.
Download the guide →