Meta’s New AI Can Turn Instagram Photos Into Deepfakes. Most Public Users Are Included Unless They Opt Out.
Meta is facing criticism over Muse Image, a new image-generation tool that can use public Instagram photos to create AI-altered pictures of real people unless users change their settings.
The tool is available through the Meta AI app and web browser, as well as on WhatsApp and Instagram Stories for U.S. users, the BBC noted. It lets users create pictures from written prompts and can pull from public Instagram accounts when a username is tagged.
Meta says users can opt out even if their Instagram account remains public. To do so, users must open Instagram settings, go to “Sharing and Reuse” and turn off the setting that allows people to reuse their content on Instagram and with AI features at Meta.
However, the feature has drawn criticism because public users are included by default. Donald Campbell, advocacy director at tech justice nonprofit Foxglove, told the BBC the tool was an “obvious recipe for disaster,” citing the spread of non-consensual AI-altered images across social platforms.
Privacy International also criticized the rollout, saying it showed how AI companies treat people’s images and data as material to be used without clear consent. The feature applies to public accounts belonging to users over 18, according to the outlet.
Muse Image was introduced this week as Meta’s first image-generation model from Meta Superintelligence Labs. The company said the tool can create and edit pictures, blend multiple photos and use presets or suggested prompts, Meta said in its announcement.
The company also said Muse Image uses invisible watermarking called Content Seal to help identify images made with Meta AI. Critics say watermarking does not address the central concern that images can be created from a person’s public profile without a direct notification.
Instagram users are not notified when their content is used to make AI-generated images, Wired reported, citing Instagram’s help center. The outlet also reported that turning off the setting or making an account private does not delete images that have already been created.
The backlash comes during broader concern over AI image tools and deepfakes. In January, U.K. regulator Ofcom opened an investigation into X after reports that Grok had been used to create and share sexualized AI images of real people, including children, Ofcom said.
AI-generated images have also been linked to misinformation during conflicts. Fake images and videos circulated during the Israel-Hamas war, raising concerns over how synthetic visuals can inflame public anger and make it harder to verify online content, Associated Press reported.
Experts also warned that Meta’s opt-out model places the burden on users. Andrew Lensen, a senior lecturer in AI at Te Herenga Waka Victoria University, told RNZ that making people opt out was not ethically acceptable because many users may not know the setting exists.
The concerns are especially serious for creators, influencers and public-facing users, whose likeness can carry commercial value. Prachir Singh, senior analyst at Counterpoint Research, told PTI that the change alters what it means to have a public Instagram account because strangers can use someone’s face to create new images without asking permission first.
