Mia Goth Slams Instagram’s ‘Incomprehensible’ Decision to Let AI-Altered Selfie Spread Online

Mia Goth did not post a selfie of herself to Instagram wearing a shirt that read “MY EX IS MY BIGGEST FAN” just days after news broke about her filing for divorce from Shia LaBeouf — and she has slammed Meta for allowing the AI-altered image to circulate. The image posted on Thursday by “miagoth_,” complete with a blue verified checkmark on the account, circulated throughout multiple news outlets as though it were a genuine post from Goth two days after her divorce filing. But a spokesperson for Goth, Christine Tripicchio, said in a statement to The Verge it could not have been genuine because, well, Goth does not have any social media accounts. Instead, Goth’s team has filed multiple complaints with Instagram over the account’s repeated posting of “counterfeit, AI images” of the actress. “The fact that this fan account meets Instagram’s criteria to qualify for a ‘blue check’ is disconcerting on a fundamental level,” Tripicchio said in the statement. “Beyond that, it is incomprehensible that Instagram evidently prioritizes the dissemination of forged and manipulated images over the protection and integrity of a public figure’s likeness.” Popular on Variety Meta and Tripicchio did not respond to immediate requests for comment. The “miagoth_” account was labeled as a fan account, according to The Verge, but its inclusion of a blue checkmark and its roughly 1 million follower count run the risk of confusing fans about the account’s veracity. While Meta does not ban AI-generated depictions of real people — instead, it labels content modified using AI as such, while also noting if a profile is an “AI-generated profile” — the account appears to have been taken down as of Tuesday afternoon. It comes as several Hollywood figures, including Scarlett Johansson, Cate Blanchett and Steven Soderbergh, have condemned deepfakes brought on by the age of AI. Blanchett cofounded the nonprofit Human Consent Foundation to help people register how AI models can — or cannot — scrape their likenesses in an effort to combat such spoofs.