Meta launches Muse Image, the AI ​​for photos in Instagram and WhatsApp

The main characteristic of the infrastructure lies in its agent-defined nature. According to what was declared by the management of the division, entrusted last year to Alexandr Wang, Muse Image does not operate in isolation but collaborates directly with the Muse Spark linguistic model. This integrated approach allows the system to analyze the text prompt, search the network and plan the composition before proceeding to the actual graphics processing. The company’s plans also include the next release of Muse Video, a model oriented towards the production of moving content designed to guarantee high temporal coherence and visual fidelity with respect to the indications provided.

From a functionality point of view, the integration introduces options for direct interaction with social profiles. By mentioning an Instagram account in the prompts, the system is able to draw on the public images of the indicated profile to reproduce its appearance in the results, while still leaving it up to individual users to manage privacy and authorizations for the use of their own content. Practical applications include transforming images using text guidelines, creating graphics for invitations or cards, and redesigning interiors based on images from Facebook Marketplace or the web.

Furthermore, the system allows you to intervene directly on the photographs through manual modifications before sharing in stories or chats. The initial rollout will enable thirty new AI-powered effects for Instagram Stories in the US market. The global distribution plan envisages a progressive extension to other geographical areas and widespread integration in all remaining sections of the Meta ecosystem in the subsequent period.