Meta’s new Muse Image model, unveiled Tuesday, can generate AI images using public Instagram posts. The catch is how it starts: public Instagram accounts are opted in by default, and the feature lets other users reuse posts, reels, and profile photos, including your profile picture, unless you change a specific setting.

In practice, Instagram profiles now automatically allow users to share and modify other people’s public posts. That includes downloading your posts for “reuse,” then remixing them with AI. Even worse for creators who value control, Instagram says existing AI-generated images made with your content won’t be removed, and users won’t be notified if their content is used by others.

So where do you stop it? There is a toggle, but it is buried. Users can switch permissions on and off only through the Instagram app, under the “Sharing and reuse” tab in the settings menu. In that menu, users can disable separate toggles for posts and reels, but the controls are not available outside the app.

This setup matters because it flips the usual privacy posture from “opt in” to “opt out.” If your Instagram account is public, your photos are potentially “fair game” for other people’s AI creations. That includes profile pictures, which are high-leverage identity assets. The feature is not limited to feed posts either; it extends to reels and profile photos, which means more surfaces of your brand and presence can be reused in AI image generation.

Meta’s intent is also pretty clear from the product positioning. Muse Image is part of Meta’s broader push to compete in generative AI, and the rollout aims to make AI image creation a built-in feature for Instagram’s billions of users. In Business Insider’s framing, Meta is moving into a crowded competitive field, rolling Muse Image to compete with image-generation tools from OpenAI, Google, Midjourney, and Adobe by using the distribution and content graph Instagram already has.

That “built-in” strategy is where second-order risk starts showing up for executives, boards, and anyone thinking about user trust as a growth lever. A default opt-in for public content is the kind of policy that triggers privacy scrutiny because it gives users less control up front. For years, Meta has faced backlash over corporate and user-facing data practices, including criticism over using public posts to train AI models by default and requiring users to opt out rather than opt in. Privacy advocates have argued that these policies leave users with too little control over how their content is repurposed.

Regulators and watchdogs typically do not need to prove “malice” to get leverage. They focus on control and consent mechanics: what users see, how defaults work, whether users can predict downstream uses, and how hard it is to reverse. Here, the reversal depends on digging through the app settings, under “Sharing and reuse.” And even if a user turns it off later, the source notes that existing AI-generated images made with their content will not be removed.

That means the real operational question for Meta and for peers is not just “can users change a toggle?” It is “what is the cost of misunderstanding and what is the duration of impact?” The lack of notifications compounds that timeline problem. If users are not notified when their content is used, they cannot easily enforce their preferences after the fact. For any platform trying to scale generative AI, this becomes a repeatable trust challenge: the product can be powerful and still create governance debt.

For leadership teams across social platforms and creator economies, the Muse Image rollout is a live case study. It shows how default permission models, identity-related assets like profile pictures, and notification gaps can turn a consumer feature into a privacy flashpoint. If you build where people’s faces and brands live, you are not just launching an AI model. You are setting the rules for who gets to remix, how long the remix lasts, and how quickly the original owner can learn about it.