Meta has rolled back image generation based on public Instagram profiles following user criticism.

Meta has hastily disabled one of the key features of its new image generation model, Muse Image — the ability to create content by referencing public Instagram accounts. The decision came just three days after launch, indicating serious miscalculations in assessing user expectations.

Muse Image, introduced on July 7 as the first image generation model from Meta Superintelligence Labs integrated into Meta AI, initially allowed users to add a public Instagram profile name to a prompt. The system then used open photos from that account to create a new image. However, by July 10, the company updated its announcement, stating that the feature "did not meet user expectations" regarding privacy.

Privacy as the Main Stumbling Block

Criticism of Muse Image was fierce. The main complaints concerned the automatic opt-in nature of the feature, which created serious privacy risks. Actress Hannah Einbinder publicly called on users to disable this option. The Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) also issued a strong statement, calling any mechanism for using images unacceptable without explicit and conspicuous user consent.

After the feature was disabled, the union supported Meta's decision, noting that any other use besides explicit consent is a "complete miscalculation of public opinion regarding the obvious dangers and harms inherent in such use."

What Remains in Muse Image

Despite the removal of the controversial feature, the Muse Image model itself remains available as a product in Meta AI. Users can still generate and edit images based on text prompts and photos. Moreover, Meta confirmed that Muse Image is already being used in over 30 new AI effects for Instagram Stories and in WhatsApp chats with Meta AI in a limited number of countries.

For context, on June 9, Meta also introduced the multimodal model Muse Spark 1.1, focused on agent tasks and coding.

Expert Commentary: This incident clearly demonstrates how fine the line is between innovation and privacy violation in the era of generative AI. Meta, possessing a vast array of user data, clearly underestimated the public backlash. Although the company responded quickly, this case will become an important precedent for regulating the use of public data in training and operating AI models. In the future, we will likely see stricter requirements for explicit user consent, which will benefit the entire ecosystem.