Forensic science has long relied on fingerprint analysis to catch suspects, but a recent study using an artificial intelligence model has revealed that fingerprints from the same person can sometimes resemble one another. Published in the journal Science Advances, the study sparked widespread controversy after editors previously rejected it, believing it undermined an “established truth” in forensic science.
Article contents:
- Historical background of fingerprints
- The role of artificial intelligence in the discovery
- Journal rejections and challenging beliefs
- Reasons for fingerprint similarities
- Implications for the world of criminal investigation
- Next steps
- Frequently asked questions
Historical background of fingerprints
Fingerprints have remained a reliable tool in solving crimes for over a century, considered unique codes that cannot be repeated. However, a team led by Hod Lipson at Columbia University discovered underlying similarities between different fingerprints from the same hand.
The role of artificial intelligence in the discovery
Undergraduate student Gabe Guo worked within this team, using a deep learning network on a sample of 60,000 fingerprints. The algorithm received images of fingerprints from different fingers of the same person and managed to detect similarities with an accuracy rate reaching 77% in some cases.
Journal rejections and challenging beliefs
The research faced difficulties in getting published; a leading forensic medicine journal refused to accept it, arguing that “every fingerprint has absolute individuality.” But the team insisted on the accuracy of the results and published them in the journal Science Advances, igniting debate over whether criminal cases that relied solely on fingerprints should be reconsidered.
Reasons for fingerprint similarities
Researchers pointed out that the algorithm did not rely on the usual minutiae pattern, but instead focused on the angles of curves and swirls in the fingerprint ridges. Consequently, “similarity families” can exist between fingers, especially among individuals with the same genetic background.
Implications for the world of criminal investigation
This discovery could lead to the reopening of criminal cases to verify fingerprint evidence, especially if it was relied upon solely without other supporting evidence. It may also force law enforcement agencies to adopt more precise verification systems that include multiple biometric checks.
Next steps
The team plans to expand the sample size and test the persistence of similarities across different fingers of multiple individuals. If the results are confirmed on a broader scale, judicial authorities may demand radical updates to the currently implemented fingerprint recognition system.
Frequently asked questions
1. What is the most prominent result of the study?
The algorithm found similarities in different fingerprints of the same person with 77% accuracy.
2. How did scientific journals handle the paper?
A specialized forensic medicine journal initially rejected it, and it was later accepted in Science Advances after thorough reviews.
3. Does this discovery negate the value of forensic fingerprints?
No, but it draws attention to the necessity of combining them with other evidence and developing more comprehensive verification techniques.
4. What is new about the analysis method?
Artificial intelligence relied on ridge angles and swirls instead of traditional endpoints and bifurcations.
5. Will old criminal cases be reopened?
This may happen in certain prominent cases if fingerprints were the sole relied-upon evidence for conviction.
6. Is the similarity restricted only to fingers of the same hand?
The study proved similarity between different fingers of the same person, but there may also be partial similarity among different individuals.
7. How do forensic experts benefit from this knowledge now?
By developing additional examination techniques and increasing awareness of the possibility of fingerprint similarities when making judicial decisions.
8. What is the next step for the research team?
Expanding the database and conducting more comprehensive experiments to confirm the pattern and the extent of its global prevalence.
