Meta has decided to disable a feature in its new Muse Image model that allowed generating images based on public Instagram profiles. This option was removed just three days after launch — such a swift reaction speaks to the scale of negative community backlash.
Let me remind you that on July 7, Meta introduced Muse Image as the first image generation model from Meta Superintelligence Labs, integrated into Meta AI. The key "feature" was the ability to specify any public Instagram account in the prompt so that the neural network could use open photos to create new images. It sounded promising, but in practice, it turned into a ticking time bomb for privacy.
Privacy Proved More Important Than Innovation
By July 10, the company officially updated its announcement, acknowledging that the feature "did not meet user expectations regarding privacy." This is a mild phrasing for a situation that sparked a wave of criticism. The problem was compounded by the fact that the option was enabled automatically, without explicit consent from account owners.
The Screen Actors Guild SAG-AFTRA issued a strong statement, calling any mechanism that uses images without explicit consent a "complete miscalculation of public sentiment." Actress Hannah Einbinder publicly urged users to disable this feature, only adding fuel to the fire of public outrage.
After the controversial feature was disabled, SAG-AFTRA supported Meta's decision, emphasizing: "Anything other than explicit and conspicuous consent for such uses of Instagram users' images is unacceptable." This is a clear signal to the entire industry: the era of mindlessly using public data for AI is coming to an end.
What Remains in Muse Image?
The Muse Image model itself as a product has not gone anywhere. It is still available in Meta AI for generating and editing images based on text prompts and uploaded photos. Moreover, Meta is already using it in 30 new AI effects for Instagram Stories and in WhatsApp chats with Meta AI in several countries. Abandoning the controversial feature is not a rejection of the technology, but a course correction.
By the way, in early June, Meta also introduced the multimodal model Muse Spark 1.1 for agent tasks and coding, so the company is clearly betting on an AI ecosystem, but now with an eye on user privacy.
My analysis: This incident is a classic example of how a tech giant overestimated audience loyalty. Meta clearly expected Instagram users to be flattered by the opportunity to become a "source of inspiration" for AI, but instead faced a wave of outrage. The market has made it clear: even public data is not free raw material for training and generation without explicit consent. I think in the coming months, we will see stricter privacy policies across all Meta products, and this is the right trend.