Facial recognition has been secretly used in universities

A Canadian university is set to remove several vending machines from campus after a student found evidence the machines used facial recognition technology.

Smart vending machines at the University of Waterloo gained attention earlier this month Reddit user SquidKid47 He shared the photo. The photo showed an M&M brand machine with an error code: “Invenda.Vending. FacialRecognition.App.exe – Application Error”.

The post sparked speculation among some online users and caught the attention of a University of Waterloo student, who told the tech news website. Ars Technica He was identified as River Stanley, who wrote in the local student magazine MathNEWS writes Stanley researched the Smart Vending Machines and discovered that they were offered by Ataria Vending Services and manufactured by the Inventa Group. Canadian publication CTV News Mars, the owner of M&M's, reportedly owns these machines.

In response to the student magazine report, the head of technology at Ataria Vending Services told MathNEWS, “It is impossible to identify a person using the technology on the machines.”

“It is important to understand that ATMs do not take or store photographs or images, and an individual cannot be identified using the technology in ATMs,” the statement said. The technology acts as a motion sensor that detects faces, so the machine knows when to activate the purchase interface – no images of customers are captured or stored.

The report says the machines are “fully GDPR compliant”. EU General Data Protection Regulation related to The regulation is part of EU data protection laws that govern how companies can collect citizens' data.

“At the University of Waterloo, Ataria is responsible for providing last-mile services – we handle inventory replenishment and snack vending machine logistics. Ataria does not collect data about its users and does not have access to identifying users of these M&M machines,” the statement said.

Invenda Group told MathNews that the technology does not store information on “permanent storage media” and that the machines are GDPR compliant.

“The software performs local processing of digital image maps received from a USB optical sensor in real time without storing this data on permanent storage media or sending it to the cloud over the Internet.”

“Only final data such as an individual's presence, estimated age and estimated gender are collected without any connection to an individual,” says Invenda Group's FAQ, according to MathNEWS.

Petersborough/Getty Images

Amid speculation, the University of Waterloo told CTV News it plans to remove the devices from campus.

“The university has requested that these devices be removed from the campus as soon as possible. In the meantime, we have asked for the software to be deactivated,” University of Waterloo representative Rebecca Elming told the broadcaster.

Representatives for the University of Waterloo, Inventa Group, Ataria Vending Services and Mars did not respond to Business Insider's request for comment.

Facial recognition technology on college campuses is a constant point of contention for students and staff, with examples occurring around the world. In May 2018 started A school China has begun monitoring students in classrooms using facial recognition technology, which is scanned every 30 seconds. Two years later, a woman said she was on TikTok Fails a test after an AI system accuses her of cheatinghad meaning.

According to The Guardian, in March 2020, tensions rose when students protested against facial recognition at dozens of US universities. reported.

“Education is supposed to be a safe place, but this technology is harming the most vulnerable people in society,” one DePaul University student told the newspaper.