Meta disables Instagram AI feature that allowed users to generate deepfakes of public accounts
Meta has removed an Instagram AI tool that allowed users to generate images referencing public accounts following concerns over potential misuse and non-consensual deepfakes.
1. Feature Removal
Meta has disabled an Instagram feature that allowed users to generate AI images by @-mentioning public accounts. The tool, which was part of the company's new Muse Image AI model, was designed to reference content from public profiles to create new AI-generated imagery. Meta stated that the feature was intended to be a creative tool, but acknowledged that it "missed the mark" following significant public feedback.
2. Public and Professional Backlash
The feature faced immediate criticism for allowing the use of public content without the explicit permission of account owners. Critics, including the National Center on Sexual Exploitation, argued that the tool could be exploited for malicious purposes such as sextortion and the creation of non-consensual deepfakes. Additionally, the Screen Actors Guild advised its members to opt out of the feature, criticizing the company for placing the burden on individuals to navigate complex settings to protect their likenesses.
3. Company Response
While Meta initially provided an opt-out mechanism for users, the company ultimately decided to remove the functionality entirely in response to the widespread concerns. In an update to its blog post regarding the Muse Image AI model, Meta confirmed that the ability to reference public Instagram accounts via mentions is no longer available.