Privacy SOS

Feds spending billions to implement face recognition technology that failed to identify Boston Marathon suspects

The FBI is spending billions of dollars to construct a ‘Next Generation Identification’ biometrics system, but recent events show that the technology isn’t ready for primetime application.

Back in April 2013, the city of Boston was on edge. Someone had blown up the finish line at the Boston Marathon on Monday the 15th, and two days later, on Wednesday April 17, police still hadn’t announced whether or not they had any suspects. Reports filed days later show they were making progress. Acting on a tip from a man who lost a leg in the bombing, and using data from the many private and public surveillance cameras that blanket the area on Boylston street, investigators zeroed in on the man they called ‘Suspect 2’. That was Wednesday afternoon. 

WaPo reports:

“[F]acial-recognition software did not identify the men in the ball caps. The technology came up empty even though both Tsarnaevs’ images exist in official databases: Dzhokhar had a Massachusetts driver’s license; the brothers had legally immigrated; and Tamerlan had been the subject of some FBI investigation.” Tamerlan had also been arrested in Massachusetts on a 2009 domestic violence charge; his photograph was therefore in the FBI’s own CJIS database.

The expensive, potentially privacy destroying biometrics tool failed to work when it counted the most. Now, Iowa republican Senator Chuck Grassley suspects the FBI is hiding something about what it knew, and when. 

In a letter to the bureau, the Boston Herald writes, Grassley asked “what feds did to attempt to ID the Tsarnaevs before publishing their photos, and why FBI agents were spotted conducting surveillance near Central Square on April 18 — just hours before authorities say the brothers ambushed and murdered MIT campus cop Sean Collier.” Grassley’s letter questions whether the FBI knew who the Tsarnaevs were before their images were released to the public.

The FBI strongly denies prior knowledge of the suspects. In a statement, the bureau wrote: “To be absolutely clear: No one was surveilling the Tsarnaevs and they were not identified until after the shootout. Any claims to the contrary are false,” [emphasis mine].

It might be true that the FBI didn’t know who the Tsarnaevs were until after the photographs were released to the public, but prior reporting quoting federal officials suggests that the bureau’s above statement is false. Again, the Washington Post: 

Once the photos of the men in caps were made public Thursday, the FBI tip line filled with calls, including one from the brothers’ aunt, who provided her nephews’ identity, according to federal law enforcement officials.

Either the FBI didn't know the Tsarnaevs' identities until 'after the shooting,' or they got a tip from the suspects' aunt on Thursday evening. Both of those things cannot be true.

Regarding the mysterious appearance of JTTF agents near the site of the shooting death of MIT police officer Sean Collier just hours before he was killed, the FBI says its people were “at M.I.T., located in Cambridge, MA, on April 18, 2013, on a matter unrelated to the Tsarnaev brothers.” The younger Tsarnaev is being charged with the murder of the MIT officer, although little to no evidence has emerged connecting him or his brother to the killing.

Senator Grassley told the Herald that his office has not yet received a response to his letter from the FBI. Perhaps the letter will do more to explain the strange coincidence that placed FBI agents at the scene of the Collier murder hours before it happened, and fill in the details about how investigators attempted — and failed — to identify the Tsarnaevs using face recognition technology. 

That latter question has implications for everyone in the United States.

The government is spending billions of dollars to construct a giant biometrics warehouse, and to implement face recognition technology at state and local police departments nationwide. But when the technology could have been most useful to the people of Boston, it failed.

In September, Boston Police commissioner Ed Davis gave a speech to UMass Lowell students. He said that in a free society, police fundamentally cannot ensure perfect security. “Everybody asks me what we could do to stop this. Do we need more cameras? Is there a computer system? Is there a piece of software to identify suspects? None of those things would work,” Davis said.

He’s right. The billions of dollars spent on so-called ‘counterterrorism’ preparedness in the United States since 9/11 didn’t do squat to stop the Boston Marathon attacks. And while the money spent on rescue training helped first responders coordinate and execute a truly impressive operation to save lives after the attack, privacy-invasive technologies like face recognition utterly failed to help investigators identify their suspects. Old fashioned police work, like interviewing the survivors of the attack, pointed them in the right direction.

In addition to providing answers to Senator Grassley's questions about what it knew, and when, the FBI should explain why it is spending billions to implement biometrics technology that clearly doesn't work.

© 2021 ACLU of Massachusetts.