Baltimore magazine|March 2021
The police "spy plane" experiment is over, but the growing surveillance of Baltimore continues.
J.Cavanaugh Simpson

DEBORAH KATZ LEVI , head of the Baltimore City Public Defender Special Litigation Section, knew something did not square. One of her clients, Kerron Andrews, had been charged with attempted murder after he was picked from a photo array of possible suspects following the drug-related shooting of three people in Southwest Baltimore in late April 2014. The Baltimore police subsequently issued an arrest warrant for Andrews and, three days later, found him inside an apartment, which wasn’t his home, just west of Leakin Park. Andrews just didn’t see how the police knew where he was. “Something kept happening to his phone, his phone kept dying,” Levi recalls. She sensed the police or state prosecutors had not disclosed everything in the case. Turns out, she was right.

Baltimore police had secretly deployed an active cell-site simulator, frequently referred to as a stingray, which masquerades as a legitimate cellphone tower and tricks mobile phones within a certain radius into connecting to the briefcase-sized device. The cellsite simulator BPD used had forced Andrews’ phone into transmitting signals that enabled police to locate its precise whereabouts. This part is critical: The simulator doesn't wait for outgoing calls to be made. It actively transmits a signal that travels through walls and reaches into cellphones to conduct a general search of cellphones within the device’s radius.

Initially, Det. Michael Spinnato had been unable to locate Andrews, whom he believed to be in hiding. But he was able to confirm Andrews’ cell number through a confidential informant and then went to court to get Sprint, Andrews’ cell provider, to turn over GPS data from his phone. That information narrowed the search down to the 5000 block of Clifton Avenue and an area of 30 or so apartment units situated around a U-shaped sidewalk. At that point, BPD’s advanced-technical team arrived and, using a cell-site simulator by the brand name “Hailstorm,” pinged the real-time location of Andrews’ phone inside a specific apartment. Spinnato rapped on the door. With the consent of the woman who answered, Spinnato and other officers entered and found Andrews sitting on the couch. His cellphone, it would later come out in court documents, was in his pants pocket. Police secured the apartment, got a search warrant for the premises, and found a gun between the sofa cushions.

In the Circuit Court for Baltimore City, however, Andrews’ lawyers successfully argued the court approval for police to access his Sprint records did not extend to use of the cell-site simulator. The judge ruled use of the Hailstorm device without a warrant was an illegal search under the Fourth Amendment and suppressed all evidence gathered from the apartment as “fruit of the poisonous tree.” It did not help that police failed to disclose the use of the Hailstorm device during the discovery portion of the trial, or that Andrews, currently incarcerated on unrelated charges, had not been picked out from two previous photo arrays of potential suspects.

More disturbing: What came to light as part of another case around that same time was the Baltimore police—which had received instruction from and signed a nondisclosure agreement with the FBI—had long been engaged in the questionable practice of using the phony cell-tower devices to surveil Baltimoreans. In 2015, a Baltimore detective testified that the department had deployed the technology 4,300 times during the previous eight years without telling defense attorneys, judges, the public, and apparently, any elected officials. In the process, the BPD had temporarily interrupted the cellular service of untold thousands of completely innocent Baltimoreans, who would’ve had no clue why their phone wasn’t working.

“[Baltimore’s police department] makes exceptionally heavy use of the equipment— BPD may even make greater use of its cellsite simulator equipment than any other city, state, or local law enforcement agency in the country,” Laura Moy, a lawyer with Georgetown University’s Institute for Public Representation, wrote in a follow-up FCC complaint. “BPD uses the equipment not only to investigate violent crimes of the most troubling nature, but also to investigate everyday street crimes, to locate witnesses, and for other unspecified purposes.”

If you reside in Baltimore City, you’ve seen and heard the police helicopter known as Foxtrot swoop down, flashing its megawatt spotlights.

You likely caught glimpses of the city’s first in-the-nation aerial surveillance experiment last year—Cessna “spy planes,” equipped with 12 cameras, flying in circles, capturing images of people in Baltimore streets and backyards, an effort also initially deployed by police in 2016 without public or elected-official knowledge. If you lived in the mid-and late 2000s in neighborhoods designated as “high-crime,” a legal term of art used to justify all kinds of police behavior, then you’re no doubt familiar with the surreal blue-light camera boxes once perched atop street poles. Prone to technical problems, they’ve been replaced over time by more than 800 closed-circuit cameras (CCTVs), currently connected to the Baltimore Police Department’s Watch Center. Those law-enforcement surveillance systems barely scratch the surface. PostFreddie Gray’s death, the BPD has upped its surveillance game. And diversified it.

For now, the “spy planes” are gone. The city canceled the contract in early February. But police-linked surveillance in Baltimore now includes Ring and Nest Hello security doorbell cameras, some of which are linked to the ever-expanding CitiWatch network of public, private, and business cameras. Automated license-plate readers can track a “hot list” of “known violent-crime offenders of investigative interest.” Meanwhile, BPD’s Axon 3 body-worn cameras, meant to help reduce violent interactions with residents, are increasingly equipped with RoboCop-like sensory surveillance capabilities. And though not widely publicized, Baltimore police officers utilize facial-recognition technology, with images submitted for comparison to a secretive Maryland state imagery database that contains all the driver’s license photos in the state. That database is decried by critics as “one of the most invasive” facial-recognition systems in the nation. And then there’s ShotSpotter, an acoustic system that detects and locates firearm discharges. It’s another private-public policing partnership, similar to the 2020 “spy plane” program which had been run by for-profit, Ohio-based Persistent Surveillance Systems.

Baltimore’s police department is not alone, of course, in using these constitutionally gray gadgets. In one Annapolis case, police used a stingray to try to track down someone who robbed a pizza delivery person of 15 chicken wings and three subs. In Detroit, ICE used a cell-site simulator to locate and detain an undocumented immigrant. In Miami, police apparently purchased a stingray specifically to surveil political protestors. As privacy implications mount—algorithm-driven “predictive policing” is already here, and drones are likely on the way—Baltimore’s next-gen spy tech scandals raise a pressing question: Why is Charm City often the laboratory for emerging surveillance technology?

It turns out, we check all the prerequisites for a 21st-century American surveillance state: high crime rates, a lack of police oversight—the Mayor and City Council have little control over BPD policies (the department has been a state agency since around the Civil War)—strained community-police relations, low-income, segregated areas, and most importantly, a history of racially biased policing. In fact, new surveillance technology—even after the scathing 2016 DOJ report that found Baltimore’s police department routinely engaged in unconstitutional practices and racially biased policing—is being deployed primarily, and sometimes solely, in predominantly African-American neighborhoods.

Take, for example, Baltimore’s public-housing complexes, 98 percent of which are considered minority households. More than 180 of the city’s 832 cameras peer from those brick apartment buildings, keeping a steady eye on roughly 14,300 people. With Baltimore’s population hovering around 600,000, that means more than a fifth of city police cameras surveil 0.02 percent of total residents, almost all of whom are Black or Brown.

Or consider last year’s six-month “spy plane” experiment. Under the contract, the planes were supposed to cover nearly the whole city. Yet an astonishing “99 percent of the company’s flights centered on East and West Baltimore—predominantly Black areas,” according to a study from the New York University School of Law’s Policing Project and publicly available flight data. The city’s northern, predominantly white, wealthier neighborhoods, where residents are also concerned about crime, were recorded on only two occasions, the report showed. Late in the pilot program, which had trouble securing a third plane, flight paths were expanded from 32 square miles to 45, partly to address the disparity. Meanwhile, other technologies, such as ShotSpotter, are deployed in the same majority Black sections of the city, five miles each in East and West Baltimore.

“[Police need to] legitimately investigate and prosecute violent criminals, and we all want safer communities,” says Stephen Smith, a retired federal judge at Stanford’s Center for Internet and Society, and a scholar in law and 21st-century police techniques. “[But] is surveillance truly done evenhandedly, or are we tilting it in such a way that it’s really used to surveil a minority community and let everybody else go scot-free?”

There’s no doubt that Baltimore police face a daunting task in solving spiraling violent crime in a city where homicides have jumped more than 60 percent since 2014. So why not tap new tools? Back when fingerprinting was first used, privacy concerns arose. In the early 1990s, newspaper op-eds questioned the legitimacy of DNA evidence. Emerging tech might prove helpful.

Baltimore’s embrace of public safety-oriented surveillance started with 16 cameras some 25 years ago. Faced by rising crime rates in the mid-1990s, BPD used a $75,000 federal grant and another $58,000 from the Downtown Partnership of Baltimore to put up cameras outside Lexington Market. Police officers manned street kiosks to monitor video footage. The 16 cameras, it was hoped, would reduce notable nuisance crimes, thefts from cars and the like. From the beginning, the Video Patrol Program, as it was called, was viewed as a pilot project.

Battle lines between civil libertarians, police, and political leaders were quickly drawn. Some critics predicted the cameras would simply shift crime elsewhere, and the ACLU stressed “we need to tread carefully so we don’t get carried away in the future.” Others believed the camera might catch a suspect here or there, though one local defense attorney cautioned persistent police camera surveillance of the public square was “a very Orwellian step.”

Then-Mayor Kurt Schmoke dismissed concerns, noting people were already being watched in banks and many retail stores by the ’90s. “Some people get nervous of Big Brother, but I think everyone wants us to make Baltimore safe,” he said. (In a memorable moment, as CBS’s Dan Rather interviewed Schmoke about the project’s “success,” someone smashed the windows of the 48 Hours truck parked 30 feet from a police kiosk. When the program aired, it noted that no arrest was made.)

Continue reading your story on the app

Continue reading your story in the magazine