The latest
Meta Platforms launched a new image-generation model called Muse Image inside its Meta AI chatbot, Bloomberg reported. It is the company’s first image model since Meta spent billions rebuilding its AI operation under Chief AI Officer Alexandr Wang.
Details
• The model: Muse Image can generate pictures from text prompts or alter existing images based on user instructions.
• App rollout: Meta said the tool will be embedded across its apps, including Instagram and WhatsApp.
• Friends and creators: Users will be able to create images featuring friends or creators using publicly available Instagram posts.
• Opt-out controls: People who do not want their content reused or remixed by AI can disable that option in settings.
• Advertiser access: Meta plans to give advertisers access to Muse Image to create marketing materials.
• Watermarking: Images created with Muse Image will include an invisible watermark.
• Safety measures: Meta said the system includes safeguards to block content that violates its terms of service, including child sexual abuse material.
Background
The release is part of Meta’s broader AI push. The company has rebuilt its AI lab as Meta Superintelligence Labs, launched a large language model in April and is developing a video-generation model expected in the coming months.
Between the lines
Muse Image is more than a consumer feature. It fits Meta’s larger plan to turn its AI models into products inside its apps, then eventually sell model access to outside developers through a cloud offering.
What to watch
Watch how quickly Meta expands Muse Image across its apps, how users respond to AI reuse of public Instagram content, when advertisers gain access, and whether Meta’s coming video model can close the gap with rivals.
