Meta has introduced a new artificial intelligence image-generation technology that allows users to modify images from public Instagram accounts. This feature has drawn criticism from privacy advocates and tech analysts. Users can edit images, including profile pictures, unless they adjust a specific setting in the Instagram app. Thorin Klosowski, an analyst from the Electronic Frontier Foundation, stated that the setting should be opt-in for users, as it represents a new use of publicly posted photos. Even if users disable AI image generation, previously generated images will remain accessible.
The Muse image model, developed by Meta Superintelligence Labs, was released recently and generates images based on social context from Instagram. Meta claims that Muse Image is their most advanced model, designed to follow instructions and edit images precisely. Alongside Muse image, Meta also previewed Muse video and announced an upgraded version of its reasoning engine, Muse Spark 1.1, as part of its efforts to enhance its AI capabilities and compete with companies like Anthropic and OpenAI.
Concerns about user safety and privacy have been raised, with Creative Artists Agency stating that creators should have control over how their likenesses are used. Previously, Meta relied on third-party AI models for image generation. The Muse image tool is now available in the Meta app, on Instagram stories in the U.S., and on WhatsApp in select countries, with plans for broader availability soon.
Additionally, Meta is investing significantly in AI infrastructure, expecting to spend up to $145 billion this year, and has secured supply agreements with companies like Samsung Electronics. However, CEO Mark Zuckerberg noted that the development of their Agentic AI technology is progressing slower than anticipated.