Meta's Muse Image Sparks Privacy Backlash Over Non-Consensual AI Photo Generation
Key Takeaways
- ▸Muse Image uses advanced AI to generate photos by blending multiple images and interpreting complex text prompts, with both free and premium subscription tiers
- ▸The tool can leverage public Instagram profile pictures to create AI-generated content without explicit user notification or consent, though an opt-out mechanism exists in settings
- ▸Privacy advocates warn the feature repeats harms already seen on social platforms and exploits user data as "raw material," raising concerns about non-consensual deepfakes and image manipulation
Summary
Meta has launched Muse Image, a new AI image generation tool available through Instagram, WhatsApp, Meta AI app, and web browsers that creates pictures from text prompts and blends multiple photos together. The tool can generate images using people's public Instagram profile pictures without explicit consent, though Meta provides an opt-out setting buried in privacy menus. The feature has drawn sharp criticism from privacy advocates and tech justice organizations who argue it enables non-consensual image manipulation similar to deepfakes, calling it an "obvious recipe for disaster." Critics note this comes as regulators worldwide—including Ofcom investigating X's Grok tool—increasingly scrutinize AI-generated imagery of real people used without permission.
- Meta plans to expand Muse Image to Facebook, Messenger, and advertising tools, with a video-generation version reportedly in development
Editorial Opinion
Meta's launch of Muse Image exemplifies tech's persistent blindness to privacy principles in pursuit of AI features. While an opt-out exists, burying it deep in settings signals Meta expects most users won't find it—treating consent as a technical hurdle rather than a right. As regulators worldwide crack down on non-consensual AI imagery, Meta's decision to build this capability directly into Instagram feels deliberately provocative, prioritizing feature velocity over the documented harms its own platforms enable.
