Meta is expanding its generative AI toolkit with Muse Image, a new tool that transforms text prompts into pictures and edits existing photos directly inside its popular apps. The system was built by the company’s Superintelligence Labs under the internal code name Mango and began rolling out on July 7. It is available at no cost through the Meta AI app, inside Instagram Stories, and within WhatsApp, though heavy users will eventually hit a cap that requires a paid subscription.
How the Image Generator Works
Muse Image accepts plain-language descriptions to produce visuals and can modify photographs in a variety of practical ways. Demonstrated use cases include erasing a distracting object from a background, placing a person in front of a landmark, or generating a functioning QR code. Users who lack a specific idea can browse ready-made prompt templates. For Marketplace shoppers, the tool can digitally insert a used piece of furniture into their own living room before they commit to a purchase. Meta is also adding a fresh set of AI-powered effects and filters to Instagram Stories alongside the launch.
A Controversial Tag-and-Generate Feature
One capability in particular is generating scrutiny. If an Instagram profile is public, another user can mention that account with the @ symbol in the Meta AI app and pull the person’s photos into a new AI image. According to Meta, a tagged username is enough for the system to assemble a finished composition from publicly available posts. The company’s policy states plainly that others can create content with your Instagram material via these AI features, and you will not receive a notification. After the scope of the feature was first reported, criticism spread quickly, with some observers calling it a privacy time bomb that incorporates real people into synthetic images without explicit consent—a practice that adds to long-standing concerns about AI-generated content flooding the platform.
How to Control the Setting
For public profiles, the permission is switched on by default, meaning users must actively opt out. The controls reside in Instagram’s settings under a section dedicated to sharing and reusing content, where individuals can disable the ability for others to use their posts and Reels with Meta’s AI features. Separate toggles exist for posts and for Reels, and Meta provides detailed instructions on its own help page. Two limitations remain: AI images that have already been generated are not removed when the setting is turned off, and the toggle only applies prospectively. The feature is still rolling out, so the switch may not be visible to every user immediately. Private profiles are inherently exempt from Muse Image access.
What Comes Next
Meta intends to extend Muse Image to Facebook and Messenger soon, with rollout to advertisers and agencies planned for the coming weeks. A video-generation sibling, Muse Video, is already under development. The privacy mechanics land against the backdrop of Meta’s past regulatory penalties, including a then-record $5 billion FTC fine in 2019 over the Cambridge Analytica matter and the decision to disable Facebook’s facial recognition system in 2021, factors that continue to color the public conversation around each new AI release.
Sources: about.fb.com, help.instagram.com, www.theverge.com