AI discovered: Fingerprints may not be unique

The latest research, led by engineering student Gabe Guo of Columbia University, raises questions about the reliability of traditional criminal investigation methods. The findings could potentially help solve unsolved criminal cases and even exonerate innocent people.

For a long time, investigators relied on fingerprints to identify suspects and link them to crime scenes. This method has led to countless arrests and convictions based on the assumption that fingerprints are unique to each individual finger and no two are alike.

However, a new study by Gay Guo using artificial intelligence has revealed something amazing. Analyzing 60,000 pairs of fingerprints from a government database, the researchers found several matches. It turned out that the prints of different fingers on the same hand can be similar, and sometimes even the same.

Artificial intelligence was able to determine with 77% accuracy which pairs of fingerprints belong to the same person, which surprised the researchers themselves. This discovery could have major implications in the field of forensic science.

Gabe Guo noted that their tool may not be perfect for identifying evidence in forensic cases, but is suitable for generating key insights in forensic investigations. This discovery could change the way we think about the use of fingerprints in criminal investigations and have far-reaching implications for the judicial system.

Experts believe that such results can help uncover new facts in criminal cases and warn against erroneous accusations.