Artificial Intelligence in Police Work

26 February 2018

Predicting and preventing crime with Artificial Intelligence – this is no longer science fiction in police work. And most Germans are in favour of such a scenario. Six out of ten German citizens (61 percent) think it is reasonable to anticipate offences with the aid of AI, so that a more effective police presence can be deployed to locations where the potential for crime is greater. This is the result of a representative survey commissioned by the German digital association Bitkom. “Artificial Intelligence can become a best friend and helper for the police and investigating authorities”, says Bitkom security expert Teresa Ritter. “With AI, crimes such as burglaries can be analysed for certain patterns, which in turn may be used to calculate the likelihood that such crimes might be committed in given areas and at given times.”

The German citizens would support the use of AI to combat financial crimes more strongly still – that is to say in identifying patterns of suspicious streams of money and preventing money laundering. Nine out of ten (92 percent) think this is a viable course of action. Almost two thirds of those surveyed (64 percent) support intelligent facial recognition, which analyses video recordings in railway stations or other public places in search of serious offenders, for example. Three out of ten residents of the German Federal Republic  (31 percent) could see the advantages of using intelligent robots to assist with police work. “Support for the use of AI in police work is consistent across all age groups. It is essential to place strict constraints on such a sensitive area, and to prevent any misuse of the technology with a closely harmonised system of checks and balances”, continues Ritter. “An AI system  cannot replace a human police officer. But it can help the authorities to protect people more effectively.”

Note on methodology: Data is based on a survey conducted by Bitkom Research on behalf of Bitkom. In the survey, 1,008 German residents aged 14 years and older were polled in telephone interviews. The survey is representative. The question was presented as follows: “Please tell me whether you think the use of Artificial Intelligence is very appropriate, quite appropriate, rather inappropriate or completely inappropriate in the following fields of application.”