University of Chicago create near-perfect crime detection AI

Category: AI Americas
Posted by AIBC Group

Researchers at the University of Chicago have discovered an AI model that, after analyzing historical crime data from 2014 to 2016, managed to predict the crime levels for the following weeks with 90 percent accuracy. This experiment was replicated in seven other U.S cities with similar levels of success.

Crime-detection AI: The good, the bad and the concerning

Published in Nature Human Behavior, these sets of studies not only focused on predicting crime but also provided valuable data on crime patterns and even self-reflective responses to crime measures.

Professor James Evans, Co-Director of MACSS and co-author of the research paper, told Science Daily that this initiative allowed the researchers

to ask novel questions, and lets us evaluate police action in new ways.

This being said, the research also shed light on the fact that socio-economic factors had a tangible effect on police action. Assistant professor at the University of Chicago, Ishanu Chattopadhyay, told Insider that the data gained from the AI models detected that crime committed in higher-income neighborhoods led to more arrests than those in lower-income neighborhoods.

Such predictions enable us to study perturbations of crime patterns that suggest that the response to increased crime is biased by neighborhood socio-economic status, draining policy resources from socio-economically disadvantaged areas, as demonstrated in eight major U.S. cities

He also told Science Daily that research had illuminated the fact that when

“you stress the system, it requires more resources to arrest more people in response to crime in a wealthy area and draws police resources away from lower socioeconomic status areas.”

Fixing the bias

In response to the crime-detection AI’s bias, Chattopadhyay noted how the researchers were attempting to reduce the effect of the bias by focusing on the sites of crime over identifying suspects. He told Insider that

“Ideally, if you can predict or pre-empt crime, the only response is not to send more officers or flood a particular community with law enforcement. If you could preempt crime, there are a host of other things that we could do to prevent such things from actually happening so no one goes to jail, and helps communities as a whole.”

Join us in Belgrade, Serbia from the 22-25th August:

With the peninsula being known for its natural beauty, rich cultural landscape and fantastic delicacies, the Balkans are renowned for having some of the most quintessentially Mediterranean vistas and experiences. From the Adriatic coastline of Split to the fairytale-esque castle of Lake Bled, the region has a lot to offer the world. This being said, something that fewer may know is the fact that the nations of the Balkans are incubating a nascent but powerfully growing technical expertise when it comes to frontier technology such as Blockchain, AI and more. Therefore the region may not only be a rich adventure into the past but my also serve as a window into the very near future.

Join us in Belgrade for the best the industry has to offer and for a window into the future of Deep Tech. To learn more about sponsorship and speaking opportunities or to inquire about attending the event, please contact Sophie at [email protected]