Not every fingerprint is unique, claims AI study

Fingerprints are less unique than we thought because a new study has found that not all of them are unique.

In a groundbreaking revelation, a recent study from Columbia University challenges the long-standing belief that every fingerprint is unique.

The study utilised an artificial intelligence program, trained to analyse 60,000 fingerprints, aiming to determine whether they belonged to the same individual. The researchers assert that the system can identify fingerprints belonging to a single person with an accuracy ranging from 75-90%, reported CNN.

Gabe Guo, an undergraduate student in Columbia's computer science program, led the research team alongside University of Buffalo professor Wenyao Xu. Published in the journal Science Advances, the study utilised a deep contrastive network, commonly employed for tasks like face recognition, to draw conclusions.

The researchers fed the AI model data from the U.S. government database, consisting of pairs of fingerprints, some from the same person but on different fingers, and others from different individuals.

Surprisingly, the study found significant similarities in fingerprints from different fingers on the same person.

The system demonstrated the ability to distinguish between fingerprints from the same person and those from different individuals, reaching an accuracy peak of 77% for a single pair, challenging the widely accepted notion of fingerprint distinctiveness.

Guo explained, "We found a rigorous explanation for why this is the case: the angles and curvatures at the centre of the fingerprint."

The researchers propose that the AI tool analysed fingerprints differently from traditional techniques, focusing on the direction of ridges in the middle of the finger rather than minutiae - specific points 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," Guo added.

Acknowledging potential biases in the data, the authors noted the necessity for more comprehensive research on a broader collection of fingerprints before implementing the technology in real forensics. The study suggests that the AI system, while potentially applicable to various racial and gender identities, requires further scrutiny.

Guo highlighted the immediate application of the findings, stating, "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."