MIAMI, FL - FEBRUARY 27:  A U.S. Customs and Border Protection officer instructs an international traveler to look into a camera as he uses facial recognition technology to screen a traveler entering the United States on February 27, 2018 at Miami International Airport in Miami, Florida.  The facility is the first in the country that is dedicated to providing expedited passport screening via facial recognition technology, which verifies a traveler's identity by matching them to the document they are presenting.  (Photo by Joe Raedle/Getty Images)
This airport commits to facial recognition tech
00:53 - Source: CNN
CNN  — 

On Friday officials said the suspect in the shooting at a local newspaper in Annapolis, Maryland, was identified using facial recognition technology.

Police had difficulty identifying the suspect using fingerprints, said Anne Arundel County Police Chief Timothy Altomare, so they turned to the state’s facial recognition software, and were able to determine the suspect was 34-year-old Jarrod Ramos.

It’s just the latest example of a law enforcement agency implementing the evolving technology in a criminal investigation. But as the use of facial recognition software becomes more common, some are asking about whether it violates citizens’ rights.

Here’s a breakdown of how the technology works, and some ethical concerns over its use by law enforcement.

How does facial recognition software work?

In the case of Ramos, authorities used the Maryland Image Repository System, or MIRS.

The Maryland Department of Public Safety and Correctional Services (DPSCS) outlined for two state legislative committees how the system worked in a December 2017 report, which the agency provided to CNN.

The report says authorities provide the software with an image of an individual they’d like to identify. The system, which is only accessible by law enforcement, then checks that image against those uploaded to the system from the state’s Motor Vehicle Administration records, the DPSCS inmate case records and mugshots, the report says.

The software compares the images to “determine the highest probability that the uploaded image may relate” to one already in the system, according to the report.

On Friday, Altamore credited MIRS with accelerating the investigative process in the hours after the shooting.

“We would have been much longer in identifying (Ramos) and being able to push forward in the investigation without that system,” Altamore said. “It was a huge win for us last night, and thus for the citizens of Anne Arundel County.”

Other agencies are using similar technology

Maryland authorities aren’t alone. The FBI has two programs that use facial recognition software, according to a statement from the FBI’s Deputy Assistant Director of the Criminal Justice Information Services Division given to the House Committee on Oversight and Government Reform last year.

One, known as the the Next Generation Identification System (NGI), uses what’s called a “repository” of mugshots known as the Interstate Photo System.

The agency provides the system with a photo of a potential suspect, and it returns a “gallery” of photos that agents manually review to see if one matches their suspect.

The other program, the Facial Analysis, Comparison, and Evaluation Services Unit (FACE), is even more comprehensive – it can search 411 million images of faces, according to a 2016 report from the Government Accountability Office.

FACE uses not only mugshots, but photos in state and federal databases, as well – including the Department of State’s database of passport and visa applications.

Like the NGI, a “probe” photo is checked against the databases, which give back a series of “candidates” that FACE checks for a potential match, the FBI’s statement said.

Critics have raised privacy concerns

Although MIRS proved paramount to quickly identifying the suspect in Thursday’s shooting, growing use of the technology has raised concerns with civil liberties groups.

The FBI stressed that both programs are used simply as an “investigative lead,” and not as a means of positively identifying suspects. The bureau also said the “probe” photos used to search software are all “collected pursuant to applicable legal authorities” over the course of an investigation.

The same goes for the MIRS in Maryland, according to the state’s DPS report. In both systems, the “probe” photos law enforcement provide are not saved. is a new startup in Hong Kong that has devised a way of harnessing Facebook "likes" as a way of getting valuable donations to worthy causes.
Does Facebook's new recognition software go too far?
04:05 - Source: CNN

Just last month, a number of civil rights organizations signed a letter to Amazon CEO Jeff Bezos asking the company to stop selling its own facial recognition technology to local police, saying it posed a threat to minority communities and immigrants.

According to emails obtained by the ACLU through a Freedom of Information Act request, local law enforcement in Washington County, Oregon, was comparing images from its jail booking photo database against images of suspects on security cameras or photos provided by citizens.

“People should be free to walk down the street without being watched by the government,” the letter, made public by the ACLU, said. “Facial recognition in American communities threatens this freedom.”

And it’s not just law enforcement agencies using it. The ACLU this month also condemned the use of facial recognition technology by US Customs and Border Patrol at the Orlando International Airport.