Report outlines biometric tools used to solve federal crimes
The U.S. Government Accountability Office has issued the first in a pair of reports sketching the digital forensic tools, including biometric systems, that federal law enforcement agencies are using or experimenting with.
The material is informative, but the results are not surprising when it comes to the use of forensic algorithms to recognize faces, fingerprints, irises and voices. While the executive branch of the government has not advertised the arsenal at its disposal, agency representatives have readily answered media questions about it.
The GAO is Congress’ independent, nonpartisan agency, tasked with knowing — or being able to find out — how taxes are spent by the federal government.
Lawmakers in two House committees (Science, Space, and Technology and Oversight and Reform) asked the office for an assessment of biometric systems at play. Part of the GAO’s goal with the new report is to show lawmakers how automation is helping to make investigations speedier and more objective.
Phase I of the report was released this month. It examines dominant algorithms being used and how they are applied to solving federal crimes. At an undefined point in the future, the second part will be released with policy options for overcoming challenges to how agents apply algorithms.
Researchers working on the first half specifically looked at the Commerce Department’s National Institute of Standards and Technology (NIST), the Department of Justice, Homeland Security and the Department of Defense.
They found that these departments generally use probabilistic genotyping, latent print analysis and facial recognition to answer the question, “Does this evidence implicate a specific suspect?” Digital analysis of irises, speech and handwriting also are used but not as much.
Probabilistic genotyping is statistical analysis of biological material against a known DNA sample.
The FBI, part of the Justice Department, uses genotyping and latent print algorithms to see if a person can be tied to a crime. It uses face recognition to generate suspect leads, according to the report.