Astronomical Techniques Unveil New Method to Detect AI Deepfakes

Researchers use eye reflection analysis to identify inconsistencies in AI-generated images, enhancing deepfake detection.

New study from the University of Hull applies astronomy tools to spot deepfakes through eye reflections.
The Gini index and CAS system measure light distribution in eye reflections to determine image authenticity.
Inconsistencies in reflections between eyes indicate AI-generated images with about 70 percent accuracy.
The method is not foolproof but adds to the arsenal of techniques to detect manipulated photos.
Researchers highlight the ongoing 'arms race' between deepfake creators and detection technologies.