An image showing actress Mia Goth in a T-shirt that appeared to comment on her divorce from Shia LaBeouf spread across Instagram and other social platforms. It looked like a personal post, but Goth’s representative says it was AI-generated and came from a fan account that had been reported to Instagram.
The episode highlights how a familiar platform badge and a convincing image can make fabricated material seem authentic, even when an account is labeled as fan-run.
How the image gained traction
After news broke that Goth had filed for divorce from actor and filmmaker Shia LaBeouf, an Instagram account called miagoth_ posted an image of her wearing a shirt with the words “MY EX IS MY BIGGEST FAN.” The post also included Beyoncé’s song “Single Ladies (Put a Ring on It).”
The account had a blue verified checkmark and more than a million followers. Hundreds of thousands of people liked the image, including people in the entertainment industry. Outlets including OK! and People wrote about it as though it were authentic, and screenshots then circulated on other platforms.
That reach gave the image a life beyond its original post. Once screenshots move between services, viewers may encounter the claim without seeing the account’s fan-run label or having a clear way to assess where the image came from.
Goth’s representative says the image was fabricated
Christine Tripicchio, a representative for Goth, told The Verge by email that Goth does not own or use social media accounts. The representative said multiple complaints had been filed with Instagram asking it to remove the verified fan account, which continued to post counterfeit AI images of the actress.
Tripicchio called the account’s blue check disconcerting and said it was “incomprehensible” that Instagram appeared to prioritize the spread of forged and manipulated images over protecting a public figure’s likeness. The complaint connects two issues: the use of AI to depict a real person and the platform signals that can make a post seem credible.
Meta did not immediately respond to a request for comment. As of Tuesday morning, the miagoth_ account appeared to have been wiped from Instagram.
What a blue check now communicates
For years, the blue checkmark signaled that an account was official and authentic. Meta later began selling blue checkmarks through a new program, which allowed accounts such as fan pages to obtain badges even when they were not run by the person shown in the content.
The miagoth_ page was identified as fan-run, but the badge could still be confusing. A viewer might see the checkmark and a large follower count before noticing, or understanding, the label. The incident shows why the meaning of a verification mark matters: people use it as a cue when deciding whether a post attributed to a public figure is real.
In this case, the image was not presented simply as a detached AI illustration. Its shirt, music choice and timing made it appear to carry a personal message from Goth. Those details helped the post fit the surrounding news, while the badge could add an appearance of authority.
AI labels leave a gap for still images
Meta’s approach is to label some generative AI content rather than ban AI-generated images of real people. Photorealistic AI videos of people may require an AI label under the company’s rules, but still images are carved out from that requirement.
That distinction matters when a fabricated still image can travel widely and be mistaken for a real post. AI-generated content featuring real people, including public figures, regularly attracts millions of views on Instagram. Meta executives have also signaled that AI-generated content will play a key role in how users engage with its platforms.
Entertainers have raised concerns about their likenesses being used to make AI content. In 2024, Scarlett Johansson said OpenAI had used a soundalike for the voice of ChatGPT after she declined to work with the company on the project. That same year, Taylor Swift said AI-generated images of her posted by Donald Trump led her to realize she needed to “combat misinformation” about who she would vote for.
The Goth image offers a concrete example of the challenge: a synthetic picture can be shared as news about a person, while a badge and the absence of a visible label may make its origin harder to judge. The central question is how platforms can help viewers distinguish a real statement from an invented one before it has spread.