Predictive Policing: Technology and the Police State

At first blush, it seems like a scene out of an 80’s era sci-fi film.  The police chief plugs into the network.  Pressing a few keys, he scans a screen full of data and radios a squad car: “Proceed to the corner of Wilshire and Bundy.  The system predicts that a robbery is about to take place.”

Somewhere else, in the plush and trendy offices of a big data company, a supervisor is called to scroll through flagged user data.  After a quick read, he verifies the messages indicate a sexual predator – and with the push of a button, sends them over to the state police department for preemptive investigation.

Neither of these scenarios are science fiction.  The software described in the first paragraph is called PredPol, while the second scene takes place at none other than Facebook Headquarters.

PredPol (an abbreviation of “predictive policing”) is a software program developed by a startup company which recently raised a round of 1.3M, and is currently employed in multiple cities across America.  Using algorithms similar to those that Amazon relies on to predict customer purchasing behavior, PredPol uses crime history mappings and frequencies to identify areas where crimes are likely to occur again.

Early results are in – and they are impressive.  PredPol is helpful not only in driving down crime (Los Angeles has seen burglaries decline by 19%, Santa Cruz by 30%) but also in allocating limited resources.  Police officers can’t be everywhere at once, but it has been demonstrated time and time again that simply patrolling problem areas helps drive down crime – software like PredPol helps determine where the “feet on the street” will make the biggest impact.

But for all the comfort of reducing crime, legal scholars are discomfited by the long-term implications of turning traditionally high-judgment areas like policing over to software.  Andrew Guthrie Ferguson, a professor at the University of the District of Columbia, warns about the risks of discrimination, over-reliance on algorithms particularly when underlying conditions have changed, and the clash with Fourth Amendment rights – freedom from unwarranted search and seizure:

[Predictive policing may]… cause courts to rethink the current overly flexible approach to reasonable suspicion, based on a concern that this technology could be manipulated or used in a discriminatory manner.

Other nations have taken the use of technology to maintain order one step further: during the Olympics, China ran surveillance-enabled taxis which allowed the police or service center to shut off the vehicle’s power if they believed the driver to be in danger.

And yet, it is Facebook’s foray into policing, that is, according to technology writer and scholar Evgeny Morozov, even more alarming.  On one hand, Facebook’s tracking of messages is hardly voluntary – US law requires that companies that serve minors provide oversight.   But, Morozov argues, putting enforcement into the hands of private data companies puts democracy at risk.  This type of “solutionist” technology seems harmless, but the price we pay is the abandonment of the checks and balances that are the foundation of the American democracy.

Finally, no curation about the melding of science fiction and policing would be complete without a link to Ray Bradbury’s Fahrenheit 451.  The iconic science fiction writer passed away last year, at the age of ninety-one.

Image credit: Jay Phagan via flickr