From Doja Cat to Lady Gaga, hyper-realistic AI images of celebrities in outfits they never wore are becoming a Met Gala tradition – and the technology making them is only getting better.
When image generation technology first started becoming more mainstream, a few AI-generated images of imagined Met Gala outfits would pop up here and there. But as the technology has become more widely available, fake images of people who weren’t even at the event have become a plague.
Every year, around the Met Gala, several platforms flood with hyper-realistic but fake AI images of celebrities in dramatic, on-theme outfits, often better or wilder than reality.
This year, posts pretending to show Doja Cat, Lady Gaga, Demi Lovato and a handful of other celebrities in elaborate looks garnered significant engagement before people called out the confusing use of AI.
Why this happens every year
According to a 2024 article by Time, the Met Gala is perfect bait. Its status as the biggest, most theatrical red carpet event builds spectator expectations for avant-garde, impractical, artistic fashion.
AI images exploit this by generating exactly what fans wish they saw – over-the-top, flawless, theme-perfect creations. Real outfits can look tame or messy by comparison, so the fakes often “outserve” reality in virality.
A look at few of the AI-Generated images at this year’s Met Gala:
Demi Lovato in an imagined Met Gala outfit. Picture: X, @MOTODEMI
Demi Lovato in an imagined Met Gala outfit. Picture: X, @MOTODEMI
Doja Cat in an imagined Met Gala outfit. Picture: X, @doctorTeeTee
Doja Cat in an imagined Met Gala outfit. Picture: X, @doctorTeeTee
Lady Gaga in an imagined Met Gala outfit. Picture: X, @ThaiJanuario
Lady Gaga in an imagined Met Gala outfit. Picture: X, @ThaiJanuario
The advancement of tools like Midjourney, Flux, Grok and Gemini’s Nano Banana let anyone type a prompt and get photorealistic results in seconds.
This does not require skills and allows people hiding behind anonymous accounts to churn these out for “clout” and sometimes even earn money from the engagement generated by these posts. This tactic is called engagement farming as posts like this rack up likes, views, quotes, and replies.
The interaction drives the algorithm and even when debunked (with several comments noting “this is AI”), the original often gets millions of impressions first.
While most of it is simply harmless fan fiction via images, what starts out as trolling, meme culture and a low-hanging fruit for accounts chasing easy growth can snow-ball into something more serious.
How to combat the misinformation fatigue
AI lowers the barrier for visual storytelling and blurs reality, eliciting the usual complaints about “slop” and misinformation fatigue.
While the phenomenon is not new, AI models keep improving, making fakes harder to spot at a glance. Social media users now spend more time pouring over images for the tell-tale signs: hands, backgrounds, lighting inconsistencies, or tattoo inconsistencies.
If this kind of thing bothers you, the easiest way to verify the veracity of an image is to check replies or reverse image search where there aren’t any easy-to-spot labels that the image was AI-generated.
Support Local Journalism
Add The Citizen as a Preferred Source on Google and follow us on Google News to see more of our trusted reporting in Google News and Top Stories.
