Meta has introduced a new image generation model, Muse Image, which uses photos and videos from public Instagram profiles as references. A key feature is the automatic shift to an opt-out model, where all users with public accounts are included in the feature without their prior consent.
What Happened
Meta launched Muse Image, which allows the use of Instagram content to create new AI images through image-to-image and video-to-image mechanisms. All users with public profiles were automatically included in this process. To prevent their content from being used, users must manually change their settings under Settings and activity -> Sharing and reuse -> Posts and videos.
Context
Traditionally, the use of user data requires explicit consent (opt-in); however, Meta is moving toward an opt-out model, where data usage is active by default for all public profiles. This allows the company to instantly scale datasets for its generative models by leveraging a massive array of existing visual content.
Why It Matters for the Industry
This move sets a new industrial precedent for aggressive data grabbing by Big Tech. It creates a powerful barrier for competitors by limiting their access to high-quality visual content, but simultaneously increases the risk of regulatory pressure and could lead to internet fragmentation as content creators migrate to closed platforms to protect their intellectual property.
Why It Matters for Users
Users' public photos and videos can now serve as the basis for generating new images and transferring visual styles by third parties without a direct request. Users need to urgently check their Instagram privacy settings to prevent the uncontrolled use of their appearance or style by neural networks.
What Is Not Yet Known / Limitations
Discussions range from the technical aspects of data pipeline implementation to serious ethical-legal risks and legal disputes regarding the use of digital likeness and style.
Sources
Author
Look at AI, Editorial Team