What to Do About the Inevitable Danger of Police Camera Surveillance?

​Police have access to more than 71,000 surveillance cameras in New York City, and to more than 40,000 cameras in Los Angeles.
 
This technology is rapidly becoming ubiquitous from coast to coast. As it does, civil libertarians are shifting from outright opposition to public surveillance cameras – which increasingly seems futile – to advocating for policy guardrails that protect privacy.
 
That American cities are going the way of London, where police cameras are on every street corner, is undeniable. The Harvard Crimson reports that Cambridge, Massachusetts, is one of the latest cities to debate whether to allow police to deploy a network of surveillance cameras. The Cambridge Police Department was on the verge of installing highly visible cameras that would surveil the city’s major parks and even Harvard Yard when the city council suspended a vote after hearing from a prominent civil rights attorney.
 
Even then, Emiliano Falcon-Morano of the Technology for Liberty Program at the Massachusetts ACLU seemed to bow to the inevitability of cameras. He recommended that this technology not be installed until the “police department addresses several important questions and concerns to ensure that it is deployed in a manner that conforms with civil rights and civil liberties.”
 
In Philadelphia Dana Bazelon, a former criminal defense attorney and frequent critic of police intrusions into privacy, is now advocating the expansion of surveillance cameras. As an advisor to the Philadelphia district attorney, Bazelon sees police cameras as the only way to stem gun violence. This turnabout prompted Reason’s J.D. Tuccille to accuse Bazelon of discarding “concerns about government abuses to endorse a wide-reaching surveillance state.” Tuccille notes how much easier surveillance cameras may make the job of policing. He archly quotes Charlton Heston’s character in Touch of Evil, “A policeman’s job is only easy in a police state.”
 
The argument in favor of public surveillance cameras is that when we step into the public square, we can expect to lose a degree of privacy. After all, no law keeps an officer on patrol from glancing our way. What’s so bad about being seen by that same officer through a lens?
 
The answer, simply, is that camera networks do more than see. They record and transform faces into data. That data, combined with facial recognition software, with cellsite simulators that record our movements by tracking our cellphone location histories, with social media posts that log our political views, religious beliefs, and personal lives, brings us to within spitting distance of a police state.
 
It is out of this concern that the Electronic Frontier Foundation has helpfully provided Americans with the ability to see the surveillance mechanisms unfolding in their communities through its Street Level Surveillance website. Yet, whether we like it or not (and we don’t like it), ubiquitous camera surveillance by the police in almost every city is coming.
 
It is coming because public surveillance is useful in solving so many crimes. As city leaders temporarily shelved the police surveillance proposal in Cambridge, a man in New York City was freed after serving 16 years in prison, exonerated by evidence from old surveillance footage. Arvel Marshall was railroaded in 2016 by a Brooklyn prosecutor who sat on the exonerating tape, which clearly showed someone else committing the murder for which Marshall was convicted.
 
There is no denying that, when the images are clear, surveillance footage can provide irrefutable identification of a criminal (or not, as in Marshall’s case). But the flip side is that the same technology, once it becomes networked and seamlessly integrated by AI, will give the powerful the means to track Americans with little more than a snap of the fingers or a click of the mouse – not just criminals, but protestors, political groups, journalists, and candidates.
 
As this new reality unfolds, questions emerge. How will police surveillance data be stored? How secure will it be from hackers? How long will it be kept? Will it be networked with other forms of tracking, such as our purchased digital data, and combined by AI into total personal surveillance? Will this data be used to follow not just potential terrorists but Americans with criminal records in a predictive effort at “precrime”? Should technology be deployed that anonymizes the faces of everyone on a tape, with deanonymization or unmasking only at the hands of an authorized person? Should a warrant be issued to watch a given crime or to unmask a face?
 
The terms of this new debate are changing as technology evolves at fast forward. But it is not too early to ask these questions and debate new policies, city by city, as well as in Congress.