Meta shipped a new AI imaging feature this week called Muse, and buried in the rollout is a default that should stop you mid-scroll: if your Instagram account is public, other users can feed your photos into the tool and remix your face into AI-generated images without asking you first.
You were opted in before you ever heard the product’s name.
How Muse Actually Works
The mechanics are simple, which is part of the problem. CBS News reported that Muse lets Instagram users take photos from public accounts and generate AI images with them, no approval or notification required. A user can @-tag any public account, pull in that person’s photo, and “build a visual” around their likeness, in Meta’s cheerful phrasing. The feature is live for US users through the Meta AI app and website, WhatsApp, and Instagram Stories.
There is an off switch, but you have to know it exists. According to IBTimes, public users are included by default and must dig into Instagram’s settings, find “Sharing and Reuse,” and disable the option that lets people reuse their content with Meta’s AI features. Keeping your account public while opting out is possible. Meta just made sure the path runs through a settings menu most people will never open.
The Opt-Out Trick Is the Whole Story
Focus on the design choice, because it is the why behind everything else here. Meta did not build Muse with a consent screen that asks, “Can other people remix your photos with AI?” It built the feature pre-enabled and put the burden of refusal on a billion-plus users, knowing the overwhelming majority will never find the toggle. AI researcher Andrew Lensen of Victoria University of Wellington called the opt-out approach ethically unacceptable for exactly this reason: consent that people do not know they are giving is not consent.
This is not Meta’s first run at the move. The company trained its AI models on years of public Facebook and Instagram posts under a similar you-could-have-objected theory, and it has spent two decades treating privacy defaults as a growth lever. The pattern holds because it works. Opt-in kills usage numbers. Opt-out delivers a launch-day feature that “everyone” is using, and the backlash gets processed later as a PR cost.
Who Actually Gets Hurt
The stakes are not evenly distributed. For most casual users, the risk is a weird AI image of them circulating in a group chat, creepy but survivable. For creators, influencers, and anyone whose face is their livelihood, a tool that lets strangers generate images from their likeness is a direct hit on the commercial value of their own identity. And for women and teenagers, the risk profile is darker: critics warn the tool lowers the friction for non-consensual altered imagery, the exact category of abuse that platforms have spent years promising to fight.
The deepfake conversation usually centers on politicians and celebrities. Muse quietly extends the raw-material pool to every public account on Instagram, which is to say, to the default state of the platform. It fits a broader pattern we have tracked as Meta pushes into new AI-driven products with the settings dialed for growth first.
What to Do Right Now
If you keep a public Instagram account, open Settings, go to Sharing and Reuse, and turn off content reuse for AI features. Do it for your kids’ accounts too, since public teen accounts are exactly the surface this feature exposes. If you run a creator or business account, treat the toggle as part of protecting your brand.
The bigger fix is not in your settings menu. Regulators in Europe and several US states already treat biometric likeness as protected data, and a feature that remixes faces by default is begging for a test case. The question worth asking is not whether Muse gets a consent screen eventually. It is why, in 2026, a company with Meta’s history still gets to launch without one.