The decision
Nikon said on 9 October that the winning video in its Small World in Motion competition did not comply with the rules on generative AI and that it has adjusted the rankings accordingly.
The statement is careful about what it is not saying. The ruling “is based solely on the submitted video’s eligibility under the rules outlined for this competition” and “should not be interpreted as a judgment of the entrant’s professional reputation, scientific contributions, or intent”. Nguyen Nam Nhat is now the winner, for a video of a tiny roundworm and a single-celled Dileptus, and the awards page has been updated.
What was in the video
The disputed entry was by Dr Ning Xu, an optical engineering researcher at the National University of Singapore. PetaPixel reported that it was meant to show the movement of cilia, the hair-like structures lining part of the lungs, in tissue Xu said came from a child with primary ciliary dyskinesia, a rare genetic disorder causing chronic lung, sinus and ear infections.
The objections came from the scientific community rather than from Nikon. Edward Phelps, an associate professor at Florida University, wrote on LinkedIn that “purple structures pop in and out of existence” and that “green cilia also appear from nowhere and do not match the known size”. Other scientists said the matter could have been settled if Xu had shared the raw greyscale video.
The five words at issue
Xu has said he used AI in post-processing and denied that this broke the rules. The rule states that “AI-generated videos are not permitted”.
That sentence is the whole dispute. It plainly bars generating a video, and says nothing explicit about generative models applied to a real recording after the fact — colouring structures, denoising, upscaling, interpolating frames. In microscopy that distinction is not academic: a great deal of modern image processing uses learned models, and the line between enhancing a measurement and inventing one runs through the middle of the toolchain.
Nikon’s judges concluded that this entry fell on the wrong side of it. Xu’s position is that it did not.
Why this matters beyond one prize
Scientific imaging competitions are where the public sees microscopy, and they run on the assumption that what is on screen was observed. The failure mode here is not a fabricated image passed off as real: it is a real recording processed until a reviewer could not tell which features were measured and which were produced.
That is the same problem research journals, photography contests and now a camera manufacturer’s prize are all confronting with the same inadequate instrument — a rule written before the tools existed.
What Nikon says next
The company acknowledges the rule did not survive contact with the case. Nikon says that advances in imaging and artificial intelligence “continue to evolve rapidly, creating new challenges for organizations across many fields”, and that the episode has shown the need for both Small World competitions “to revisit rules and evaluation procedures for future competition entries”.
What to watch
The rewrite. A rule that works has to say what processing is allowed rather than what generation is banned, and the obvious mechanism — requiring the raw file on submission, which is what the scientists in this case asked for — is the one Nikon has not yet committed to.
Whether next year’s rules require raw footage is the measurable outcome of this decision. Everything else is a statement of intent.
