"The head of Argus (as with stars the skies) / Was compass’d round, and wore an hundred eyes."
So writes Ovid about Argus Panoptes, the many-eyed god and sleepless watcher who sees everything. Too bad Microsoft and the New York Police Department aren’t of a literary (or ironic) bent. They missed out on a cool handle for their slightly terrifying new spy computer, which is saddled with a rather mundane name: the Domain Awareness System (DAS).
The exact capabilities of DAS are still a little sketchy right now. I requested details from Microsoft, but they were still assembling them at press time. What little we know comes from a press conference in which Mayor Michael Bloomberg crowed, "We’re not your mom-and-pop police department anymore. We are in the next century. We are leading the pack."
It was probably too much to expect that man — who thinks your salt, soda, fat and breast milk intake is of intense concern to his city — could resist ushering New York into the bold new world of 21st-century surveillance states.
And they’ll even be able to make a boatload of money at the same time. The NYPD developed the program in cooperation with Microsoft, so New York gets its system for free. If Microsoft is able to sell the system to other police departments, New York scoops up 30% of each sale. Nice work, if you can get it.
But what exactly does DAS do? Right now, it sorts through the feeds from the city’s 3,000 closed-circuit television (CCTV) cameras and compares them to databases on criminals and potential terrorists. Most of these cameras are in Lower and Midtown Manhattan (the areas south of Canal and between 30th and 60th Streets, from river to river), but the NYPD is already expanding the coverage into the boroughs.
The system also scans license plates, allowing it to "track where a car associated with a murder suspect is currently located and where it’s been over the past several days, weeks or months," according to New York Police Commissioner Ray Kelly.
The NYPD offered the following features of the new system:
- Investigators will have immediate access to information through live video feeds and instantly see suspect arrest records, 911 calls associated with the suspect, related crimes occurring in the area and more.
- Investigators can map criminal history to geospatially and chronologically reveal crime patterns; police commanders can query databases to map, review and correlate crime information with the deployment of resources
- If a suspicious package is left at a location, the NYPD can immediately tap into video feeds and quickly look back in time and see who left it there.
- If radiation detectors in the field set off alarms and alert the Lower Manhattan Security Initiative command center, the new system will help quickly identify whether the radioactive material is naturally occurring, a weapon or a harmless isotope used in medical treatments.
Let’s unpack that a bit and see just what it means from a technology perspective and from a privacy perspective.
What DAS does is centralize and streamline the surveillance cameras already in use and any that go online in the future. The system simplifies the process of accessing and sorting through a massive data stream. Raw information on individual incidents and crime trends can be analyzed more efficiently and quickly, allowing the department to put police where they’re needed. Dispatchers can access a more complete picture of an area, and investigators can call up vital surveillance information related to a criminal act. Cars can be tracked citywide by a network of license-plate scanners.
Used properly and judiciously, these abilities may indeed be a benign tool to aid law enforcement. After all, there’s no assumption of privacy when someone is in a public space. If you can be seen by a dozen bystanders on a street corner, how does it become an invasion of privacy to be seen by a few more via CCTV?
How does DAS change things? The answer is complex, and the problem arises from emerging technologies outpacing our legal understanding of their applications. There are two key areas that are of grave concern: data usage and data retention.
Let’s take automatic license-plate readers (ALPRs) as a starting point. They were intended for use in tracking stolen vehicles and suspects of specific crimes, but they soon were being used to find uninsured drivers and other minor scofflaws. As their role expands, the ability of DAS to both sort and store vast amounts of ALPR data begins to a feel a bit ominous. The state doesn’t need to place a tracker on a vehicle, which might require some level of judicial oversight. They can track any vehicle anywhere and create a complete record of a person’s every movement, stretching back years, all without that person ever being accused or suspected of wrongdoing.
The same concerns apply to surveillance cameras. Since we don’t yet know what’s under the hood of DAS, we can’t be sure just what it’s doing, but the early specs make it sound like it may include facial-recognition software. (How else is it comparing CCTV footage against criminal databases?)
Now let’s match that power up with another emerging technology. Facebook is piloting a new system (acquired last month in the purchase of Israel’s Face.com) that allows users to volunteer for a facial-recognition retail program. Whenever the Facebook user enters a store, a camera in the stores maps his face, identifies him and creates some kind of as-yet-ill-defined custom shopping experience for him. Facebook has been using Face.com’s software for about a year and a half, training it by using all those photos people voluntarily post to the site.
Are you getting a little creeped out yet?
Now, let’s put that same power in the hands of the state and imagine it being used for something more than just texting you a coupon for McDonald’s. Imagine a web of cameras that cannot only record crimes and criminals, but record, track and store the movements of law-abiding citizens.
What happens to the Fourth Amendment at that point? What happens to the presumption of privacy? Of innocence?
And what happens when political dissenters become targets?
What happens when, for example, pro-lifers are designated as potential terrorists, as Homeland Security Secretary Janet Napolitano did in 2009? What then?
There have already been troubling instances of the NYPD subjecting innocent people to unwarranted surveillance. Muslims with no criminal histories were not only investigated in New York, but the NYPD crossed beyond its jurisdiction, into New Jersey, to conduct the surveillance.
There are supposed to be all sorts of limits on the use of the DAS, and perhaps we can rely on them. But even those limits have their exceptions. Video may only be held for 30 days — or for more if the NYPD deems it necessary. Various metadata such as license-plate scans are held for five years. The as-yet-undefined "environmental data" (possibly the radiation scanners) can be held forever. The system is intended to ferret out terrorist threats, but the city concedes that the data will be used for routine law enforcement as well.
Technology changes things. It may seem like a logical step from the security guard monitoring a fuzzy camera to the NYPD monitoring every street corner in the city, but it’s a step too far — and it will not be the last.
The automation and the data retention made possible by systems like DAS create a new kind of power: the ability to identify, track and log the movements of citizens who should be living their lives under a presumption of innocence and privacy. DAS is one more step down the road towards a complete surveillance state, where every person is tagged and tracked by a cold camera eye — and every citizen is a suspect.
Thomas L. McDonald writes about religion and technology at GodandtheMachine.com.