Meta's new AI image generator, Muse Image, has sparked backlash from privacy advocates for allowing users to generate photos that incorporate faces from public Instagram profiles by default. The tool, released Tuesday, can tag public Instagram accounts and pull from images in those posts, all without notifying the individuals whose photos are used.
Default Opt-In Raises Concerns
Privacy advocates argue that the default setting for data sharing is unfair. The cybersecurity company Malwarebytes described the situation in a 9 July blogpost, noting that switching a public Instagram account to private is the most comprehensive solution, though it is a blunt one. For those who wish to remain public, a more measured approach involves adjusting settings in the "sharing and reuse" section to toggle off the option allowing others to reuse content.
A Meta spokesperson stated to the Guardian that Muse Image was built "with strong controls and safety guardrails from day one." The company emphasized that private accounts and those of users under 18 are automatically excluded, and adult users with public accounts can opt out using easy-to-use controls. Meta also said it would take action against any content violating community standards.
Opt-Out Process Criticized as Burdensome
Privacy advocates contend that expecting users to proactively disable this feature is unfair. Danny Bradbury, a writer for Malwarebytes, described finding the setting as "its own adventure." The privacy-focused company Proton published a blogpost stating that "data sharing is turned on by default, the opt-out is buried deep in settings, and public backlash becomes the main way users find out what happened to their content." They warned users to watch the toggles closely, as the on and off states look nearly identical, making it easy to leave one active by mistake.
Children's Faces at Risk
While Meta notes that Muse Image cannot tag accounts of users under 18 and that teens cannot use the feature to tag or reference other accounts, the company has not clarified whether children depicted in photos on public adult accounts could be used in prompts. Proton warned that children's faces in public photos risk being appropriated.
Thorin Klosowski, a senior security and privacy activist at the Electronic Frontier Foundation, said: "This is the sort of setting that should absolutely be opt-in for Instagram users. It's a new use of the photos they've been posting publicly for years, and certainly wasn't on anyone's mind when they signed up for Instagram years ago."
Potential for Misuse
Meta's press release for Muse Image showcases benign uses, such as researching summer fashion looks or enhancing landscape photos. However, the tool can also combine multiple images—such as a specific person, a particular bike model, and a visual style—into a single prompt, allowing for highly customized outputs. Privacy advocates warn these features could have a dark side. The Malwarebytes blogpost noted: "Meta says Muse Image is meant to make AI image generation more personal by letting people reference public Instagram accounts in their prompts. That may sound fun when you're creating images of yourself. It's less appealing when anyone else can do the same with your account."
Reporting and Future Expansion
Meta states that Muse Image includes robust protections to block policy-violating content. Users can report objectionable AI-generated images by pressing and holding the image and selecting a "thumbs down" option. The tool is currently available in the Meta AI app, Instagram Stories in the US, and WhatsApp in limited countries. Meta plans to expand its use to Facebook and offer similar video capabilities to creators and Meta AI.
