Crimes are not something new. There is evidence that criminals have been around since we have written records. What has not happened so far is that thanks to those previous records, crime can be prosecuted and even prevented. Artificial Intelligence (AI) as a tool against crime is already used in several countries around the world.
As artificial intelligence is a broad-spectrum tool — driving autonomous vehicles, helping political decision-making, or improving our health are some of its benefits. It is a rare crixqme sector that it will not pursue. It is already used to manage police resources, detect false reports, solve crimes and even advise criminals to prevent them from going back to their old ways. We take a tour of some of the AI tools used in crime prevention.
Artificial intelligence to solve crimes?
In mid-2018, the Netherlands published a program capable of rescuing the cold cases, open cases whose tracks had long since gone cold. A large number of crimes end up archived for lack of threads to pull, but it is expected that the development of artificial intelligence will be able to rescue a few. At least if police reports are digitized, of course. Most of them are still handwritten.
Also in the Netherlands, the Terre des Hommes association designed a bait for pedophiles. Her name is Sweetie and she is a fake girl who chats looking for criminals. In just 10 weeks, Sweetie was able to find 20,000 pedophiles and identify at least 1,000. Scott Robert Hanses, an Australian, was the first to be arrested. This happened in 2013. The program is being used by police around the world.
With funding from the European Union, in 2017 VALCRI (Visual Analytics for Sense-making in Criminal Intelligence Analysis) began to be developed, a program that uses artificial intelligence combined with visual analysis and closed-circuit cameras to identify suspects. At the moment, it is still in the experimentation phase, but it promises results.
The use of cameras in AI is very common because we have been training advanced algorithms with millions of images for years. Google Image Search is one such example. Another is the photographic retouching that we find in mobile phones such as the Huawei Mate20 Pro or Huawei P20 Pro, capable of selecting advanced settings at the time of the shooting.
The importance of ethics in pre-crime
Knowing when a crime will be committed is impossible unless you have reliable data to guide you in this regard. The idea is not new. In 1931, Chicago-based criminologists Clifford R Shaw and Henry D McKay began collecting data on juvenile delinquency. They concluded that location and time frame are key values when pursuing future crime.
Today, pre-crime units use big data to establish so-called “hot spots,” places where the next crime is most likely to emerge. However, what data to use is also key if you are looking for ethical systems. We have been wrong in the past.
For example, the first pre-crime systems mobilized assets to marginal neighborhoods due to the bias of the police who for years had directed their operations there. This bias has been corrected in New York with an AI-based facial recognition system that has managed to locate 21,000 identity frauds in a matter of two years.
Also in New York, they use the CompStat program, which aims to prosecute minor crimes. Human trafficking mafias, for example, use capital from equally illicit activities, such as cloning cards or robberies at ATMs. To do this, they monitor certain profiles throughout the city in an attempt to nip crime in the bud.
At the same time, in the Californian town of Santa Cruz, they use PredPol, which from historical data can determine, hour by hour, where a robbery or gang fight is most likely to take place. Something similar, although much more invasive, they do in China with programs like Watrix, just released in November 2018.
Looking at our country, in recent years a police team has managed to develop, with considerable success. A system capable of distinguishing false from true complaints by how they have been written.
A false declaration not only implies the loss of time of the agent who collects it after it a device is activated that investigates the event. Thousands of euros simply for starting the case, many more if it continues.
What if you stop crime?
In England, they have opted for a different approach to pre-crime. If China imprisons without prior notice those who believe they will commit a crime, New York mobilizes patrols to hot spots and scans documents to verify if a complaint is behind a crime, in England, they have chosen to advise criminals not to repeat offenders. Also using AI.
In this way, Artificial Intelligence (AI) has much to contribute to the creation of a more just society, and the prosecution of uncivil behavior is one of the first steps.