Every year, the camera company Nikon runs a contest for the best pictures and videos taken through a microscope. Scientists from all over enter, and the winners get seen around the world.
This year’s top prize went to a researcher named Ning Xu. His video showed tiny hair-like structures called cilia, beating in the airway of a child with a rare genetic condition. It was beautiful — and it won first place on September 15.
But within days, other scientists started spotting things in the video that didn’t look biologically possible.
Xu explained what happened: his team had used an AI tool to clean up the video in post-processing — “enhancing the visual presentation” of the area below the cilia. He said they weren’t making any claims about what those enhanced parts actually showed.
On October 9, Nikon disqualified the video. The competition’s rules say AI-generated entries aren’t allowed. Nikon was careful to say the decision was about the rules — not about Xu’s reputation or his intentions. They also said they’ll revisit the rules for future contests, because this clearly isn’t the last time it will come up.
The prize went instead to Nguyen Nam Nhat, for a video of a tiny roundworm and a single-celled organism called Dileptus.
Here’s why this matters to the rest of us. The line between “cleaning up” and “faking” is getting blurry fast — and it’s not just scientists dealing with it. Your phone does this every day. When you snap a photo, your phone quietly uses AI to sharpen faces, brighten skies, and scrub out noise before you ever see the picture. At what point is it still “your” photo?
Contest rulebooks were written for a world where editing meant cropping and adjusting color. AI changes the question entirely. Science contests will have to rewrite their rules — and honestly, the rest of us will too.