Two women walk slowly through a parking garage in Santa Cruz, California. Occasionally, they glance into the windows of the cars they pass.
Given that information alone, a police officer would arguably have no business stopping to detain and question those women. But in 2011, that's exactly what they did. The two women were arrested for drug possession and outstanding warrants, the New York Times reported.
How did this happen? A digital crime-mapping tool told those officers to look out for automobile theft in that area. The software purports to tell the future, so the cops had reasonable suspicion.
Crime mapping has been around for decades, but a new class of products peddled by Silicon Valley-style startups claims to forecast where crime will happen, and when. So if police are told that the very specific area you're walking through or living in is prime territory for a particular crime, the question is: If the police can predict crime, can they detain you before you've done anything wrong?
The Fourth Amendment was never meant to deal with pre-crime mapping. Reasonable suspicion, which allows police to stop suspects and make brief detentions, isn't a hard science. It exists in the amorphous space before the probable cause needed to actually arrest or search someone.
There are some firm rules. "Matching a description" is a famous, well-trod example of probable cause. Carrying an instrument that might be used in a crime — for instance, if the women in that parking garage were carrying bent wire hangers — counts too. So does swerving between lanes on an empty road late at night on a Saturday.
Reasonable suspicion has traditionally been based on what Andrew Guthrie Ferguson, criminal law professor and Fourth Amendment scholar, calls "small data" — information that police can observe in a single moment about a person's behavior, clothes, belongings and activity, taken together with environment-based context clues. Is this person carrying industrial tools? Are they behaving erratically, or do they appear to be casing a neighborhood (that is, looking for ways to break into houses)? This information is gathered using a police officer's own senses and judgment.
Big data, on the other hand, uses information pulled from surveillance dragnets, cameras and social media databases to try and learn more. Does the person in this neighborhood even live here? Are those tools they're carrying appropriate for his line of work? Do they have a criminal history? And, most presciently, what kind of crime is likely to occur in this area?
Does the person in this neighborhood even live here? Are those tools they're carrying appropriate for his line of work? Do they have a criminal history? And, most presciently, what kind of crime is likely to occur in this area?
Obviously, the aim of these systems is to create more effective, objective policing — to make decisions based on data and informed insights rather than a police officer's whims. Ideally, officers who can use information to get better context can do better policing, stopping people who are more likely to commit crimes and leaving the rest of us in peace. But Ferguson is skeptical of the databases police are currently working with.
"Right now we're trying to collect as much data as we can, and police are trying to draw links and connections," Ferguson told Mic. "But we haven't become sophisticated in our data collection, even if we're moving at a rapid rate in applying it our of context on the street."
Pre-crime mapping tools are mostly predicated on prior crime behavior, which paints a selective picture of crime that is already based on traditional police practices. In California, decades-old gang databases have been shown to be full of errors, built on invasive and prejudicial surveillance that catches up disconnected individuals in the dragnet. There is, undoubtedly, a human undercurrent in police data. And while these tools are still young, they can be used just as successfully as discarded methods like New York City's controversial stop and frisk practice to selectively target individuals.
"As technology moves forward, it's eviscerating the already weak reasonable-suspicion standards on the street," Ferguson told Mic. "As police have more information about identity, it becomes easier for them to justify stopping people they want to stop, even if what they're doing isn't criminal."
Police tech is getting better at surveillance, and some are scrambling to fight back. Cameras, social media databases and biometric scanning are on the way to providing officers with quick ways of identifying people on the street without their knowledge, and connected devices are bringing those tools right into squad cars.
Of course, there's a backlash against digital surveillance from individualist contrarians who use a whole utility belt of tools to undermine new policing techniques — methods that verge into what would look like tinfoil-hat paranoia to those unfamiliar with just how far contemporary tech has come.
Elizabeth Joh, a law professor at the University of California whose research focuses on police technology, calls these "privacy protests." In a paper on the Fourth Amendment implications of taking yourself off the grid, she wrote:
The police tend to think that those who evade surveillance are criminals. Yet the evasion may only be a protest against the surveillance itself. Faced with the growing surveillance capacities of the government, some people object. They buy "burners" (prepaid phones) or "freedom phones" from Asia that have had all tracking devices removed, or they hide their smartphones in ad hoc Faraday cages that block their signals. They use Tor to surf the Internet. They identify tracking devices with GPS detectors. They avoid credit cards and choose cash, prepaid debit cards or bitcoins. They burn their garbage. At the extreme end, some "live off the grid" and cut off all contact with the modern world.
It's the kind of surveillance evasion that led Daniel Rigmaiden, a tax fraudster living off the digital grid, to discover the existence of "stingrays" — a device used by dozens of federal agencies to sweep up cellphone locations by pretending to be cellphone towers. By evading police surveillance (and being arrested and prosecuted regardless) Rigmaiden was able to determine what exactly police were truly capable of by process of elimination, doing his research from jail.
But privacy protests and taking yourself off of the grid can't protect you from predictive crime mapping, which targets neighborhoods and makes you guilty by association of the space you're walking around in. Unless your privacy protest includes cloaks of invisibility, it's a type of suspicious association from which, as of now, the Fourth Amendment has no way of defending you.