Every Fingerprint Is Not Unique, Reveals New AI Study

We have always heard that each fingerprint on a person's hand is unique. However, this theory is currently being contested by a Columbia University study. Recently, an artificial intelligence program was trained by a team at a US university to analyse 60,000 fingerprints in an attempt to determine which ones belonged to the same person. According to the researchers, the system can determine if fingerprints belong to a single individual with 75-90 per cent accuracy, as per a report in CNN. 

Gabe Gua, an undergraduate student in Columbia's computer science programme, oversaw a research team on the topic, with University of Buffalo professor Wenyao Xu serving as one of his co-authors. Published this week in the journal Science Advances, the report seemingly upends a long-accepted fact about fingerprints. The scientists used an AI model known as a deep contrastive network, which is frequently used for tasks like face recognition, to arrive at the study's conclusions. After giving it their spin, the researchers fed the data from the US government database in pairs, some of which came from the same individual (but on different fingers) and others from different persons.

It was then discovered that fingerprints from different fingers on the same person had significant similarities. As a result, it was able to distinguish between fingerprints belonging to the same person and those that weren't, with a single pair's accuracy peaking at 77 per cent, seemingly refuting the idea that each fingerprint is distinctive.

Mr Guo said, "We found a rigorous explanation for why this is the case: the angles and curvatures at the centre of the fingerprint."  The researchers believe that the AI tool analysed the fingerprints differently than conventional techniques, emphasising the direction of the ridges in the middle of the finger rather than minutiae-the places where individual ridges stop and split. "They are great for fingerprint matching, but not reliable for finding correlations among fingerprints from the same person. And that's the insight we had," he added. 

The authors acknowledged that there could be biases in the data. The study added that even while they think the AI system functions mostly the same for both racial and gender identities, more thorough research of a more comprehensive collection of fingerprints is necessary before the technology can be used in real forensics.

"The most immediate application is it can help generate new leads for cold cases, where the fingerprints left at the crime scene are from different fingers than those on file. But on the flip side, this won't just help catch more criminals. This will also actually help innocent people who might not have to be unnecessarily investigated anymore. And I think that's a win for society," he stated.