Mia Goth AI Image Shows How Fan Accounts Can Blur Celebrity Identity on Instagram

A verified fan account turned an AI-generated image of Mia Goth into what looked like a personal statement about her divorce. The image showed the actor in a T-shirt reading “MY EX IS MY BIGGEST FAN,” paired with Beyoncé’s “Single Ladies (Put a Ring on It).” The apparent joke was easy to mistake for a real post, but Goth’s representative says the image was fabricated—and Goth does not use social media.

The episode points to a practical weakness in platforms’ handling of synthetic media: labels and detection systems can help, but they don’t necessarily stop a false image from circulating under an account that appears credible. In this case, the account had a verification badge and more than a million followers. Its fan-page status did not prevent an outlet from initially describing the image as Goth’s own selfie.

The post followed news that Goth, known for Pearl and MaXXXine, had filed for divorce from Shia LaBeouf. The Instagram account @miagoth_ presented the image in a context that invited readers to connect it to that news. People initially reported it as a selfie, describing the shirt as Goth’s fashion choice. Both the article and post have since been removed.

Verification can add credibility to a fake

Goth’s representative, Christine Tripicchio, told The Verge that Goth “does not have ownership of, nor does she engage in social media on any platform.” Tripicchio said her team had filed multiple complaints asking Instagram to remove the account, which she said continued to publish counterfeit AI images.

She also questioned how the fan account qualified for a blue check. A badge can be read as a signal that an account is authentic, even when its name or profile description identifies it as a fan page. The result is a mismatch between the account’s apparent credibility and the status of the content it publishes. Meta told Mashable the account had been removed for impersonating a public figure.

That response addresses the account itself. It leaves a broader product question: how should a platform distinguish fan-made material from content that misleads people about what a person said, wore, or did? The distinction matters because an AI image can be technically synthetic while still appearing plausible in a social feed. A caption, song choice, or timely reference can give it the shape of a real personal update.

Fabricated moments can serve different purposes

The Goth image fits a wider pattern of AI accounts using recognizable people to attract attention. A March investigation by Indicator found six AI-generated influencer accounts with roughly a million Instagram followers combined. They posted fabricated encounters with celebrities; one account, @ayannasoblack, shared fake gym photos with Dwayne “The Rock” Johnson and John Cena. All but one linked to Fanvue profiles selling access to synthetic adult content.

Other accounts draw on performers’ work to build audiences. AI-generated influencer accounts have reproduced comedians’ jokes and performances, according to Mashable’s reporting. Comedian Sarah Purdum told Mashable that women’s material was being “stolen and repurposed for sexy AI models.” The underlying mechanism is familiar: a recognizable face, voice, or performance can help synthetic content gain attention, while the person whose identity is used may have had no role in making or sharing it.

Fabrications can also attach people to statements or events they never took part in. In January 2024, sexually explicit AI images of Taylor Swift spread online. SAG-AFTRA called them “upsetting, harmful, and deeply concerning.” Later that year, Swift cited a separate set of fabricated images falsely suggesting she supported Donald Trump when announcing her support for Kamala Harris. “It really conjured up my fears around AI, and the dangers of spreading misinformation,” she wrote.

In 2025, Scarlett Johansson objected to an AI-generated video that depicted her and other Jewish celebrities protesting Ye’s antisemitism. Johansson condemned antisemitism while making clear she had not participated in the video. “We must call out the misuse of AI, no matter its messaging, or we risk losing a hold on reality,” she said in a statement, as reported in coverage of her objection to the video. The same problem can affect families of deceased performers: in October 2025, filmmaker Zelda Williams asked people to stop sending her AI videos of her father, Robin Williams, saying the recreations were “NOT what he’d want.”

AI labels don’t establish that an image is real

Meta’s stated approach is to leave manipulated content available unless it violates another platform rule, such as rules against bullying or harassment. Its policy explanation says it adds “AI info” labels when it detects certain technical signals or when people disclose that they used AI. That means a missing label cannot be treated as proof that an image is genuine.

The labels have changed as Meta adjusted its system. In 2024, it replaced “Made with AI” with “AI info” after some photos with minor AI retouching received the stronger label. It later moved labels for detected AI edits into a post’s menu, while keeping visible labels on content it identifies as AI-generated.

Detection remains an incomplete part of that system. Meta has acknowledged that its tools cannot identify every AI-generated image, and that invisible signals used for detection can be removed. So the label can offer useful context when it appears, but it cannot settle the question of authenticity when it does not.

For users, the practical check is to look at the account itself and seek confirmation from a person’s known channels before treating a surprising post as a direct statement. For platforms, the Goth case shows why account identity and synthetic-media labels need to work together: a convincing image can travel farther when the account around it looks credible. Removing the account limits that particular source, but the next fabricated moment may arrive with the same familiar cues.