Nikon says its Small World in Motion winner broke the generative-AI rules

Nikon has re-ranked the 2026 Small World in Motion contest. On a statement posted 9 October 2026, Nikon says that after re-evaluating the winning video and its supporting materials, and consulting members of the judging panel, it “has determined that the winning video did not comply with the competition rules regarding generative AI.” The rankings on the site have been adjusted. Nikon says the decision “is based solely on the submitted video’s eligibility under the rules outlined for this competition. It should not be interpreted as a judgment of the entrant’s professional reputation, scientific contributions, or intent.”

The statement does not name the entrant. Nature, PetaPixel, the BBC, and USA Today identify the video as Ning Xu’s: cilia, the hair-like structures that line airways, from a child with primary ciliary dyskinesia. Nature describes Xu as an optical engineer at the National University of Singapore. PetaPixel describes him the same way, as a researcher there. The BBC identifies him with Tsinghua University.

The new order

Nikon’s statement names the new first-place video: Nguyen Nam Nhat of Nguyen Nam OME in Chu Se, Gia Lai, Vietnam, “A tiny roundworm and a single-celled Dileptus explore their microscopic world,” differential interference contrast, 40X. The 2026 gallery lists him first, then Benedikt Pleyer, Andrew Moore, Patrick Hickey, and Igor Adameyko.

What scientists raised, and what Xu said

The dispute is older than the 9 October ruling. Nature’s 30 September article says Xu was named the winner on 15 September, and that on 22 September Nikon updated a blog post about the entry to say he used an “unsupervised” AI model to assist with “post-processing.”

Nature quotes Edward Phelps, a bioengineering researcher at the University of Florida: “The purple structures resemble mitochondria but such a sub-epithelial structure composed of extracellular mitochondria of the same size as cell nuclei does not occur in biology.” He said the blue structures resemble nuclei but do not behave like nuclei, and that it is unclear what the red stain is supposed to be. PetaPixel quotes him saying “purple structures pop in and out of existence” and “green cilia also appear from nowhere and do not match the known size.” The BBC’s 30 September article reports the same kind of objection, including cilia that “appear from nowhere” and a concern that they seemed too big.

On 9 October the BBC quoted Robert Hirst of the University of Leicester, lead scientist for an NHS centre for PCD diagnosis, who said the cells and cilia “look nothing like those from PCD patients.” That is Hirst’s assessment as the BBC reports it. Nikon’s statement says the ruling is not a judgment of professional reputation, scientific contributions, or intent.

Xu’s reply, as the outlets report it, is that the AI step was coloring, not the experiment. Nature says he wrote on LinkedIn that AI was not used to generate the experimental movie, the cilia, or their motion, and that “It was applied afterwards to the reconstructed grayscale data to distinguish and color structures with similar morphology.” The BBC’s 30 September article quotes that denial and his line that AI was used “to distinguish and visualize features in the reconstructed grayscale images.” USA Today, on 1 October, reports that he said the colors are false and “do not correspond to fluorescence channels identifying specific cells or organelles.” USA Today also reports a Nikon LinkedIn comment of 26 September that it was “carefully re-reviewing” the video while Xu was “respectfully complying.” On 30 September the BBC reported Nikon saying “At this time, we don’t see that any rules were violated.”

Where a contest draws the line

The rules, on Nikon’s rules page and the entry portal, say “AI generated videos are not permitted” and that the sponsor may request the original video for verification. They also say submissions generated by AI, script, macro, or other automated means are void. They do not define which post-processing steps count as generation.

PetaPixel says other scientists argued the dispute could have been settled if Xu had shared the raw grayscale video. USA Today quotes Valentin Dunsing-Eichenauer, of Der Simulierte Mensch in Berlin, calling on Xu to “immediately make the raw data available.” Nature quotes Melanie White, a developmental biologist at the University of Queensland: “Scientific images are not just illustrations. They are data, and we need to be able to trust that what we are seeing is grounded in the underlying measurement.” Markus Sauer, of the University of Würzburg, told Nature that AI visualization becomes a problem when images misrepresent, exaggerate, or alter findings.

Nikon says the episode shows that Small World and Small World in Motion need to revisit rules and evaluation procedures. Its own rules already let it request the original video, and the scientists quoted above asked for the raw grayscale frames.