Scientists from the University of Chicago in the United States have successfully developed a new artificial intelligence system capable of accurately predicting crimes before they happen. The new system features AI technologies, such as facial recognition, with hopes of reducing crime rates and providing better capabilities to expedite the arrest of criminals and suspects. However, the results have not always been highly accurate.
Ishaan Chatterjee, an assistant professor at the Division of Biological Sciences at the University of Chicago, and his colleagues created an AI model that analyzed crime data in the city of Chicago from 2014 to 2016 and then predicted crime levels in the weeks following the data input into the program.
The team was surprised to find that the AI model predicted the likelihood of certain crimes occurring throughout the city one week in advance with an accuracy of up to 90%. The system was also trained and tested on data from seven other U.S. cities, yielding a similarly impressive performance.
Additionally, the researchers used the data to investigate areas where human bias impacts police work. They found that crimes in wealthier neighborhoods resulted in more arrests compared to poorer areas, which may suggest a bias in how police respond to alerts and reports.
