“We don’t know for sure how it does it” – AI baffles scientists by correctly assigning individual fingerprints to one person, but they are unclear how it actually works


A groundbreaking study at Columbia University is challenging the long-held belief that each fingerprint on a person’s hand is entirely unique.
According to the BBC, the research team has developed an AI tool capable of analyzing human fingerprints to determine if they belong to the same individual.
The researchers report that the technology was able to identify, with an accuracy of between 75% and 90%, whether prints from different fingers originated from one person. However, the exact workings of how the AI achieved this remains a mystery.
Helping to solve crimes
“We don’t know for sure how the AI does it,” confessed Prof Hod Lipson, a roboticist at Columbia University who oversaw the study. To make sure it was correct, the team repeated the process over and over again, each time getting the same results.
The researchers’ theory is that the tool is analyzing the fingerprints – which are formed at birth – in a way that’s different from how it is done traditionally, focusing more on the orientation of the ridges at the center of a finger rather than the pattern in which individual ridges end and fork, known as minutiae.
Despite the promising results, the Columbia University team concede that further research is necessary. The AI tool was trained on a large dataset of 60,000 fingerprints, but a more substantial number, including partial and low-quality prints, would be needed to further refine the technology.
The findings of this study, which has undergone peer review and will be published in the journal Science Advances, could significantly impact biometrics and forensic science. As the BBC reports, if an unidentified thumbprint is discovered at one crime scene and an unidentified index fingerprint at another, the AI tool could potentially link the two prints to the same individual. That won’t – for now – be good enough to solve crimes on its own, but it could help the police with their investigations.
The revelation that fingerprints might not be unique to every person hasn’t come as a surprise to everyone. Graham Williams, a professor of forensic science at Hull University, admitted, “We don’t actually know that fingerprints are unique. All we can say is that as far as we are aware, no two people have yet to demonstrate the same fingerprints.”
More from TechRadar Pro
A groundbreaking study at Columbia University is challenging the long-held belief that each fingerprint on a person’s hand is entirely unique. According to the BBC, the research team has developed an AI tool capable of analyzing human fingerprints to determine if they belong to the same individual. The researchers report…
Recent Posts
- Top digital loan firm security slip-up puts data of 36 million users at risk
- Nvidia admits some early RTX 5080 cards are missing ROPs, too
- I tried ChatGPT’s Dall-E 3 image generator and these 5 tips will help you get the most from your AI creations
- Gabby Petito murder documentary sparks viewer backlash after it uses fake AI voiceover
- The quirky Alarmo clock is no longer exclusive to Nintendo’s online store
Archives
- February 2025
- January 2025
- December 2024
- November 2024
- October 2024
- September 2024
- August 2024
- July 2024
- June 2024
- May 2024
- April 2024
- March 2024
- February 2024
- January 2024
- December 2023
- November 2023
- October 2023
- September 2023
- August 2023
- July 2023
- June 2023
- May 2023
- April 2023
- March 2023
- February 2023
- January 2023
- December 2022
- November 2022
- October 2022
- September 2022
- August 2022
- July 2022
- June 2022
- May 2022
- April 2022
- March 2022
- February 2022
- January 2022
- December 2021
- November 2021
- October 2021
- September 2021
- August 2021
- July 2021
- June 2021
- May 2021
- April 2021
- March 2021
- February 2021
- January 2021
- December 2020
- November 2020
- October 2020
- September 2020
- August 2020
- July 2020
- June 2020
- May 2020
- April 2020
- March 2020
- February 2020
- January 2020
- December 2019
- November 2019
- September 2018
- October 2017
- December 2011
- August 2010