Nikon strips microscopy video of top prize after reversing its AI ruling
The company had defended the entry as permissible enhancement. A further review changed that decision, and Nikon says it must revisit future rules and judging procedures.
Loading page…
The company had defended the entry as permissible enhancement. A further review changed that decision, and Nikon says it must revisit future rules and judging procedures.
Listen to this story
The reversal leaves an unresolved practical question for AI-assisted scientific imaging: Nikon has not disclosed what its additional checks found or how future submissions will be assessed. After consulting judges and reviewing Ning Xu’s video and supporting materials, Nikon ruled the entry ineligible under Small World in Motion’s generative-AI rules; it did not make a finding about Xu’s intent or professional reputation. Nguyen Nam Nhat now leads the contest, and Nikon says both of its imaging competitions will revisit their rules and evaluation procedures.
Nhat’s winning video shows a roundworm and a single-celled organism, Dileptus, exploring their surroundings.
Nikon moved the other winners up one place: Benedikt Pleyer is second, Andrew Moore third, Patrick Hickey fourth, and Igor Adameyko fifth.
Xu said he used generative AI to distinguish and visualize features in reconstructed grayscale research footage, not to create the cilia or their motion.
Nikon’s microscopy video contest has a new first-place winner after the company disqualified Ning Xu’s entry on October 9. A further review found that the video violated its generative-AI rules, reversing Nikon’s earlier defense of the work as permissible image enhancement. The company is also preparing to revisit how future entries are evaluated.
Nguyen Nam Nhat of Vietnam now holds the top spot in Small World in Motion. His video shows a tiny roundworm and a single-celled organism called Dileptus exploring their surroundings. Nikon adjusted the competition rankings after removing Xu’s entry, moving the other winning videos up one place.
The revised results put Benedikt Pleyer second for footage of jellyfish larvae suspended in water droplets, and Andrew Moore third for synchronized division in neighboring cells. Patrick Hickey is fourth for a video of mitochondria and chloroplasts in Turtle Vine leaf cells. Igor Adameyko takes fifth with footage of a small, colorful sea slug.
Nikon had initially accepted Xu’s explanation that AI enhanced data produced through microscopy rather than generating images from scratch. As The Scientist reported, the company explicitly described that use as compliant with its competition rules. Its new decision follows weeks of controversy over whether the finished video faithfully represented the underlying biology.
The company said it re-evaluated the video and supporting materials and consulted members of the judging panel before finding it ineligible. A spokesperson told The Scientist that the additional review involved both Xu and the judges. Nikon declined to disclose the details of those checks or discussions.
This 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.
Nikon, in its statement reproduced by Ars Technica
Xu’s entry depicted cilia beating in airway tissue from a child with primary ciliary dyskinesia, a rare respiratory disorder. The controversy concerned structures visible in the processed footage, not simply the fact that software had been used. Researchers questioned whether the video’s appearance was consistent with actual cellular biology.
Ars Technica, citing the BBC’s reporting, described researchers’ concerns about structural abnormalities and cell features that appeared and disappeared in unexpected ways. Others identified what looked like AI watermarks in source files. Those observations formed part of the public challenge to the entry; they were researchers’ concerns, not a detailed account of Nikon’s final review findings.
Xu said the video began as actual research footage and that he used generative AI for postprocessing. He described AI’s role as distinguishing and visualizing features in reconstructed grayscale images. He denied using it to generate the experimental movie, the cilia themselves, or their motion—a distinction central to his defense of the entry.
University of Leicester scientist Robert Hirst challenged the result more directly. As Gizmodo reported, Hirst told the BBC that he had spent 20 years diagnosing primary ciliary dyskinesia by examining cilia movement, length, and cell size and shape. He said the structures shown looked nothing like those from patients with the condition.
Nikon said the incident showed a need for both Small World and Small World in Motion to revisit rules and evaluation procedures for future submissions. Its statement described the competitions as celebrating scientific and artistic imaging, while committing to the integrity of the images and videos selected for their winning galleries.
Loading discussion...
Join the conversation
Explain where you would draw the line between clearer images and altered evidence.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.