Meta Pulls Instagram AI Update After Everyone Said They Didn’t Want to Be Deepfaked

After intense backlash last week, Meta pulled the most troubling part of its newest AI update on Instagram. You will no longer be easily able to create deepfakes of other Instagram users.

“We’ve heard the feedback that this feature missed the mark, so it’s no longer available,” Meta wrote on Friday in an update to its announcement blog post.

Last week, in an attempt to prove that Meta’s billion-dollar leap into generative AI had purpose, the company brought its newest AI image model to Instagram. It came with a new ability to “tag” any public Instagram accounts and include those people in photorealistic AI images — essentially creating deepfakes. 

Watch this: AI Is Indistinguishable From Reality. How Do We Spot Fake Videos?

03:15

Meta offered two ways to opt out, but many were incensed that they were automatically defaulted to opting in. Actors union SAG-AFTRA, a fierce advocate for protecting artists’ likenesses from AI companies, said in a statement: “With the dangers of nonconsensual digital replicas well known to all, a feature that encouraged that behavior is unwise. We appreciate its discontinuance. It is the responsible thing to do.”

As AI image and video models advance, it’s more difficult than ever to tell with the naked eye whether something is real or created with AI. Invisible watermarks and more prominent labels for AI content are two ways tech companies are working to help us distinguish AI content, but in a sea of AI slop, it’s still difficult. AI tools, like the image creation and editing one that’s still available in Instagram, may only make the problem worse.