Instagram's Verified Celebrity Deepfake AI Images Highlight Platform Risks

Image: Illustrative image · Instagram · Public domain · Source

AI-generated celebrity images with verified Instagram accounts grow widespread, raising concerns about impersonation, misleading content, and platform safeguards.

A recent incident on Instagram revealed significant challenges posed by AI-generated deepfake images of celebrities on social media platforms. An Instagram account @miagoth_, marked with a blue verified checkmark and claiming to be actress Mia Goth, posted an AI-generated image of the actress wearing a T-shirt that read "MY EX IS MY BIGGEST FAN" shortly after she filed for divorce from actor Shia LaBeouf. Accompanied by Beyoncé's “Single Ladies (Put a Ring on It),” the post garnered nearly 500,000 likes, including from entertainment industry peers, despite the account being a fan page and inauthentic.

Goth’s representative confirmed to The Verge that they had previously reported the account to Instagram for posting counterfeit AI images. Nevertheless, Instagram’s system permitted the fan account to obtain verification status, a decision called "disconcerting" by Goth's spokesperson, who criticized the platform for prioritizing manipulated images over protecting the likeness rights of public figures.

This viral AI image not only spread on Instagram but also permeated celebrity gossip outlets and appeared prominently in Google search results, misleadingly portraying it as an authentic comment on Goth’s divorce. Instagram eventually removed the fraudulent account for impersonating a public figure.

The proliferation of AI-generated celebrity deepfakes extends beyond this case, with numerous Instagram accounts producing manipulated images and videos of well-known figures like Scarlett Johansson, Taylor Swift, and others, often sexualized or misleading. Meta, Instagram’s parent company, confirmed to The Verge that it removed several such accounts following inquiries but acknowledged ongoing challenges in policing this content.

Meta enforces policies against impersonation and mandates AI-generated videos to be labeled, but still images remain exempt from AI disclosure requirements, limiting transparency about manipulated content. This inconsistency allows AI-generated images to circulate widely without clear user awareness.

Compared to other platforms, YouTube employs a more robust likeness detection system developed with the entertainment industry, scanning for both visual and sound-alike content and accommodating takedown requests under public interest considerations. However, the sheer volume and evolving nature of AI-generated media complicate enforcement across social networks.

The Mia Goth incident illustrates broader failures in trust and safety measures on social media, where AI-generated imagery can mislead users, propagate false narratives around sensitive personal matters, and infringe on individual rights. With no comprehensive legal framework universally in place for AI likeness rights, the onus remains on platform policies and technological solutions to mitigate misuse.

As AI content generation tools become more accessible and sophisticated, social media platforms face mounting pressure to enhance verification, content labeling, and enforcement mechanisms to protect public figures and users from deceptive and harmful AI-created media.