Nobody is safe from celebrity AI deepfakes
An AI-generated image falsely showing Mia Goth wearing a pointed T-shirt message spread widely on Instagram after her divorce filing, despite the account being labelled a fan page. The episode shows how paid verification, social media reach and search results can make fabricated celebrity images appear credible and amplify them.
The image attracted nearly 500,000 likes, and gossip coverage helped it appear in Google results, including an AI-generated overview that treated the claim as real. Instagram later removed the account for impersonation, after Goth’s representative said they had complained about its counterfeit images before. The article also describes sexualised AI depictions of other female celebrities and notes that the spread of such content continues despite earlier disputes over public figures’ likenesses and voices.
- A fake Mia Goth image drew nearly 500,000 likes.
- Paid verification helped a fan account look credible.
- Celebrity deepfakes continue to spread across platforms and search.
New here? Start with this
AI tools can create convincing images of real people without their permission. Because celebrities are widely photographed, these images can be mistaken for genuine pictures or statements, even when they come from fan accounts.
Social media platforms, news and gossip coverage, and search engines can all help a false image reach more people. Paid verification marks may also make an account seem trustworthy, though they do not guarantee that its posts are authentic.
These fakes can mislead the public and use someone’s likeness in ways they have not agreed to, including sexualised depictions. The issue raises questions about how platforms and search services identify false material and respond when it spreads.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Platforms and search services should take stronger steps to identify and limit fabricated celebrity images, especially when paid verification and prominent placement can make them seem authentic. The Mia Goth episode suggests that labels alone may not prevent widespread deception, while sexualised deepfakes can cause serious harm to the people depicted; clearer provenance and quicker responses to complaints could protect both individuals and public trust.
The case against
Aggressive moderation can remove lawful fan expression, satire and artistic work, and automated systems may struggle to distinguish these from deceptive fabrications. Requiring platforms or search services to judge authenticity or likeness claims could lead to mistaken removals and give powerful public figures undue control over discussion; measures should therefore be carefully scoped and allow meaningful review and appeal.