Following last year’s deadly shooting, Michigan State University will be implementing a new AI surveillance system to increase safety on campus. 

A new story by Bridge Michigan says the system, which tracks movement, counts crowd sizes and reads license plate numbers, has experts worried about potential privacy concerns. 

University of Michigan School of Information professor Mark Ackerman says these AI-based surveillance systems are “not very accurate in low lighting” and have been shown to demonstrate implicit bias against minorities. 

“They're very problematic about seeing certain kinds of hostile intent,” he says. 

Additionally, these security systems may also be over relied on, says UMSI professor Kentaro Toyama. 

Police, he warns, can become “so used to relying on the AI that they stop being critical about whatever information they are being fed by it.”

The new system comes after a lone gunman killed three students last year, injured five others and exposed weaknesses in MSU’s security systems. 

RELATED 

Read “After shooting, MSU adding AI surveillance to detect threats, count people” at Bridge Michigan. 

Learn more about UMSI professors Mark Ackerman and Kentaro Toyama by visiting their UMSI faculty profiles. 

— Noor Hindi, UMSI public relations specialist