The policeman stops the suspect, points the gun, asks to raise their hands. The boy refuses. And, when he reaches into his bag, the officer chills him, mistakenly thinking he is about to take a weapon. Fortunately, the only real element is … the policeman. The scene is a kind of video game: an interactive system with a screen on which various scenarios appear, with criminals and suspects played by actors. The agent evaluates what happens, talks to the virtual characters and decides what to do, whether to try to calm down or shoot. Interactive simulators (such as the VirTra system) are the high tech version of routine police training: a technology to “train” you to react appropriately in difficult conditions. On the streets, then, law enforcement officers will be able to count on the support of robots such as Knightscope (above). He moves, reads car license plates, recognizes faces and records suspicious behavior, sounding the alarm. For the moment it has been tested for security in companies (see Focus n. 256), but it is designed for the surveillance of even crowded places, such as shopping malls or school campuses. The “robocop” patrol shift therefore seems to have begun: the Dubai police have even announced that, starting from 2017, they will introduce robots capable of interacting with citizens and connecting to the police headquarters in some public places. In short, even if it will be necessary to wait for the cyborgs capable of arresting criminals, the future of the police is marked by the use of new technologies. But what consequences will they entail? Will they serve to make our cities safer? And what are the risks instead?
FUTURE ROBBERY. One of the emerging technologies really looks like something out of the Minority Report movie. These are predictive software: they do not anticipate a specific crime, but indicate where and when crimes may be committed and thus direct the police. In the field, Italy is at the forefront. A program, KeyCrime, has been operating in the Milan Police Headquarters since 2007. “Studying the robberies in pharmacies”, explains the inventor, assistant chief of the Police Mario Venturi, “I realized that much of the data collected was lost and that there was a need for a tool to compare them “. Thus was born a huge database that collects all information on crimes (yesterday only robberies, today all) with an algorithm that correlates elements such as date, place, time, type of crime, objective, data and behavior of the criminal, any video of security cameras. «At this point the software compares the single crime with the previous 8,000, contained in the archive, and proposes to the agent events potentially connected with the one analyzed. Thus a seriality between several crimes is highlighted, which allows us to “predict” the following ones as regards the type of objective, an area and a time span, but also the danger of the criminal », explains Venturi. “For example, two months ago we managed to catch some robbers who were operating in pharmacies armed with sawed-off shotguns, which was rather anomalous, waiting for them in the area and in the time frame in which the software told us they would strike.”
Today it is the man who, using the computer, directs research. But in the new version of the program (KeyCrime Cube, which Venturi hopes to propose as national software to the Ministry of the Interior), the process will be even more automatic, for example in reconstructing the seriality between crimes. “Even if the last word will be man,” says Venturi. This is not a detail, given that similar programs used in other countries have sparked controversy for their “automatism”. It happened in Chicago, where an algorithm created a list of the 400 potentially most dangerous people in the city, including children with minor offenses. In addition, predictive software generally identifies “hotspots” in the city where the risk of certain crimes is highest: the police could follow analyzes that continue to point towards the poorest neighborhoods or neighborhoods full of minorities.
Not to mention that in the future the software could draw on personal data, obtained from social media, to build potential criminal profiles of every citizen. “Data collection and predictive software prepare a future with less freedom,” argues Jay Stanley of the American Civil Liberties Union, the US organization that defends civil rights, “and the same is true of increasingly used technologies, such as video camera networks ». In fact, privacy is the other big issue. It is in play or in all cases in which citizens could be filmed by robot cameras or “body cams”, which in the US are increasingly hooked to police uniforms or glasses to record what is happening around them. While the latter document the interactions between policemen and citizens (suspect or not), “rules must be established to make sure they are not used as surveillance tools,” explains Jake Laperruque, of the Center for Democracy and Technology in Washington. . “It means making sure they are turned on only when required by the service, that access to the video is controlled, and that the use of facial recognition software is limited.” In Italy, the Privacy Guarantor, as part of the experimentation underway in Turin, Milan, Rome and Naples during public events, has requested its use “only in the event of a real danger to safety”.
CONTROL YOURSELF? And in the US, a debate has also raged over the use of drones by law enforcement agencies, to survey the cities from above: the citizens of Seattle, for example, rebelled, until the mayor blocked their ‘adoption. “After September 111, 2001, in the US there has been an escalation in the use of technology for control purposes, including by local police,” Stanley points out. The landscape of security technologies is diverse. They range from ShotSpotter, a system of sensors (in use in various US cities) that identify the origin of a shot to immediately send a patrol to the place, to Explorer, a ball with a video camera inside that shoots in different directions and that agents can launch in a building to have a complete view; from Leedir, a platform and app with which citizens can send photos and videos, to NeoFace Watch by Nec, with which people captured by video surveillance cameras can be identified.
Faced with more intelligent and pervasive tools, what future awaits us? “The more technologies are in the hands of law enforcement, the more questions will arise about how their use should be limited,” explains Laperruque. So, to put limits on possible abuses, “we need new laws,” concludes Stanley. To feel safe and not “spied on”.