When a short microscopy video won Nikon’s Small World In Motion contest, it quickly became the centre of a heated discussion among scientists and contest judges. The controversy centres on whether the winning entry, submitted by Dr Ning Xu of Tsinghua University, complies with the competition’s explicit ban on generative AI.
Background
The Small World In Motion contest, run annually for the past 15 years, celebrates images and videos captured with powerful microscopes. Entrants are allowed to enhance colour and contrast, practices that are common in microscopy, but the rules forbid the use of generative AI to create new visual content.
The Winning Submission
Dr Xu’s entry is a short video that purports to show cilia—tiny, hair‑like structures—moving inside the airways of a child with a rare disease. The footage is brightly coloured, a technique that is normally achieved by staining the sample. In a LinkedIn response, Xu acknowledged using AI "to distinguish and visualize features in the reconstructed grayscale images" but insisted that AI was not used to generate the cilia or their motion.
Scientific Concerns
Several scientists have publicly questioned the authenticity of the video. Edward Phelps, an associate professor at Florida University, noted that some structures appear to "pop in and out of existence" and that the cilia seem to emerge "from nowhere." Ian Donovan, a PhD student at UT Southwestern Medical Centre, reported a distinctive watermark that he believes indicates AI generation. Two former judges also expressed doubts about the video's realism.
Contest Rules and AI
The contest’s rulebook states that participants may not use generative AI to produce images; the work must be captured directly through a microscope. Dr Patrick Hickey, who placed fifth in the 2026 contest, emphasized that the award’s value lies not only in the £2,250 prize but also in the recognition it brings to researchers who spend long hours in dark rooms. He argued that allowing AI‑enhanced content would be unfair to other entrants.
Implications for Future Contests
Nikon has announced that Dr Xu is "respectfully complying" with an ongoing investigation and has provided additional technical documentation. The company currently sees no rule violation, but it has pledged an update once the review is complete. The episode raises broader questions about the boundary between legitimate image processing and AI‑driven creation in scientific publishing.
- Transparency: Researchers may need to disclose any AI tools used during image processing.
- Guidelines: Contest organisers could consider more detailed definitions of permissible post‑processing.
- Trust: Maintaining confidence in scientific visuals may require independent verification mechanisms.
As the investigation proceeds, the scientific community watches closely, aware that the outcome could shape how microscopy images are presented—and judged—across future competitions.