Meta's Muse Image Model Generates AI Photos with Instagram Users

Meta has launched Muse, a powerful new image generation model developed by its Superintelligence Labs, marking a significant expansion of AI capabilities across its social media ecosystem. The model is now integrated into Meta AI, Instagram, and WhatsApp, allowing users to generate photorealistic images directly within these platforms. What sets Muse apart from other AI image generators is its ability to pull real Instagram users into generated photos. For example, a user could create an image of themselves and their friends at a fictional beach party, with the AI accurately rendering each person's likeness based on their public Instagram photos. This feature opens up creative possibilities for personalized content, but it also raises serious privacy concerns. Meta has implemented an opt-out system for public accounts: users who do not want their likeness used in AI-generated images must manually adjust their privacy settings. For private accounts, the feature is automatically blocked. However, critics argue that placing the burden on users to opt out is insufficient, especially given the potential for misuse, such as creating non-consensual or misleading images. The rollout is part of Meta's broader strategy to embed generative AI into everyday social interactions. On Instagram, users can access Muse through the messaging interface or the story creation tool. On WhatsApp, it appears as an option in chat, enabling friends to co-create images in real time. Meta AI, the company's virtual assistant, also gains the ability to generate images on command. Meta emphasizes that all generated images are watermarked with invisible metadata indicating they were created by AI, in line with industry standards for transparency. Additionally, the company has deployed content moderation systems to prevent the generation of harmful or explicit material. Despite these safeguards, privacy advocates are calling for stricter regulations. The Electronic Frontier Foundation has highlighted the risk of deepfake-style abuse, where someone's image could be used in compromising scenarios without their knowledge. Meta counters that the opt-out system gives users control, and that the feature is designed for positive, creative use cases. As Muse rolls out globally, it represents both the promise and peril of AI integration in social media. Users are encouraged to review their privacy settings and decide whether they want their public photos to fuel the next generation of AI-generated content.