Researchers conclude ShotSpotter is 'too unreliable for routine use'

Occurred: May 2021

Researchers at the MacArthur Justice Center concluded that the ShotSpotter gunfire detection system sent Chicago police officers on over 40,000 'dead-end deployments' and is too unreliable for routine use.

Of 46,743 ShotSpotter alerts generated July 2019-April 2021, only 5,114 resulted in officers filing a report 'likely involving a gun,' according to the study’s analysis of records obtained from city’s Office of Emergency Management and Communications. SoundThinking claims ShotSpotter is over 97 percent accurate.

The study was the basis for an amicus brief filed on behalf of Chicago-based community groups that asked a Cook County judge to question whether ShotSpotter reports should be allowed as evidence in a murder case.

The findings raised serious questions about the accuracy and reliability of the ShotSpotter system. The lack of accuracy, together with the system's use in predominantly Black and Brown communities, fed 'racialized patterns of overpolicing,' attorneys argued.

Operator: Chicago Police Department
Developer: SoundThinking/ShotSpotter
Country: USA
Sector: Govt - police
Purpose: Detect gunfire
Technology: Deep learning; Neural network; Machine learning
Issue: Accuracy/reliability; Bias/discrimination - race, ethnicity
Transparency: Governance