Hardware firm Bunnings and department store Kmart collect customers’ “faceprints” in some locations.
Consumer advocacy group Choice says the technology is unethical, invasive and being used without proper consent or reasoning.
Both retailers defended its use as an anti-theft and safety measure.
The Australian Information Commissioner said her office had opened an investigation to determine whether they had breached privacy laws.
Australian retailers can only collect sensitive biometric information if “reasonably necessary” for their operations and they have “clear consent”, Angelene Falk said.
“While deterring theft and creating a safe environment are important goals, using high privacy impact technologies in stores carries significant privacy risks,” Commissioner Falk said last month, after the use of the technology was first revealed.
“Retailers need to be able to demonstrate that it is a proportionate response.”
Last year, she found convenience store chain 7-Eleven had interfered with customers’ privacy by collecting faceprints in a similar case.
The watchdog said it was also “conducting inquiries” about another retail company, The Good Guys, which has paused its use of facial recognition technology.
The Australian Human Rights Commission has called for a ban on the technology until Australia has specific laws to regulate its use. It followed police in Western Australia using it for Covid isolation checks.
■ The nation where your ‘faceprint’ is already being tracked
■ Facebook to end use of facial recognition software
Choice said Bunnings and Kmart were only disclosing their use of the technology in small “conditions of entry” notices at the front of stores, and in privacy policies online.
The consumer group surveyed more than 1,000 households and found more than 75% had no idea the technology was in use.
“Using facial recognition technology in this way is similar to Kmart, Bunnings or The Good Guys collecting your fingerprints or DNA every time you shop,” said Choice’s Kate Bower.
Bunnings said its use of the technology had been inaccurately characterised and there were strict controls around its use.
The data collected is not used for marketing purposes, it says, and the only images retained are of people banned from stores or those suspected of illegal or threatening conduct.
“In recent years, we’ve seen an increase in the number of challenging interactions our team have had to handle in our stores and this technology is an important tool in helping us to prevent repeat abuse and threatening behaviour towards our team and customers,” said chief operating officer Simon McDowell.
A spokesperson for Kmart also said the technology was on “trial” to prevent theft and was subject to strict controls.
The number of police forces joining Ring’s partnership program across America more than doubled in 2020, despite there being little evidence that it’s an effective crimefighting device. The rise is sparking major privacy concerns.
Newly released figures indicate that nearly 2,000 police departments across the US are partnered with Amazon’s Ring, in the process expanding the reach of the highly controversial civilian surveillance network yet further.
Ring, bought by the e-commerce giant in February 2018 for a fee that could be as much as US$1.8 billion, is best-known for producing a range of ‘smart’ doorbells, which house high-definition cameras, motion sensors, microphones, and speakers.
Not long after its purchase, the partnership program was launched – under its auspices, Ring offers authorities access to video footage recorded by the millions of internet-connected devices its customers have mounted to their homes.
In turn, citizens alerted to suspicious or outright criminal activity outside their residences by Ring’s motion sensors can submit reports directly to law enforcement via an accompanying app, ‘Neighbors’.
The figures show a staggering 1,189 departments nationwide joined the program in 2020 – the total now stands at 2,014, including 305 fire departments.
Only two US states – Montana and Wyoming – aren’t home to forces enrolled in the program, which saw partnered departments collectively request videos related to over 22,335 incidents during 2020 alone.
Ring hails the initiative as a ‘Neighborhood Watch’ revolution, which makes areas safer by helping deter andsolve crimes. Its website features numerous clips of apparent criminals caught in the act on doorsteps, and a May 2019 case in which Ring footage played a pivotal role in the identification and capture of an individual who abducted an eight-year-old was well-publicised.
Conversely, several slick promotional videos in which various police departments touted Ring’s crime-fighting capabilities have since been removed from the web.
In one such segment, which focused on the company’s partnership with police in Winter Park, Florida, the local department’s chief spoke effusively about the “value” of Ring cameras “in helping us solve crimes” – as police officers “cannot be everywhere,” the force was said to “rely” on citizens using the ‘Neighbors’ app to report incidents.
The reason for the video’s deletion is unclear, although it may be related to a February 2020 NBC News investigationwhich found Winter Park authorities had in fact not made a single arrest facilitated by footage obtained from Ring cameras since it partnered with the company in April 2018.
The story was much the same elsewhere. NBC interviewed 40 law enforcement agencies in eight states that had partnered with Ring for at least three months – three said the ease with which citizens can share Ring videos meant officers wasted time reviewing clips of issues such as raccoons running around and petty disagreements between neighbors, while others noted the deluge of footage rarely led to positive identifications of suspects, let alone arrests.
A total of 13 agencies had made zero arrests as a result of Ring footage, and several others, including those in large US cities, simply didn’t know how many arrests had been made as a result of their Ring partnership, even though they’d been partnered with the company for over a year.
Concerns over the program’s law-keeping efficacy are nonetheless somewhat secondary to its grave privacy implications.
Digital rights group Electronic Frontier Foundation (EFF) has long been a fervent critic of the system, dubbing it “a perfect storm of privacy threats” and contending that Ring and comparable ‘home security’ providers serve to greatly inflate paranoia about crime, transforming every innocent delivery person, charity fundraiser, or election canvasser into a potential – if not likely – criminal with every motion sensor update beamed.
“By sending photos and alerts every time the camera detects motion or someone rings the doorbell, the app can create an illusion of a household under siege,” EFF argues. “It turns what seems like a perfectly safe neighborhood into a source of anxiety and fear. This raises the question: do you really need Ring, or have Amazon and the police misled you into thinking that you do?”
Police departments are greatly incentivized by Ring to further this feedback loop. In areas where police are partnered with the company, departments are granted credits with which they can buy more cameras to distribute to residents, for every resident who downloads the Neighbors app. As such, officers are encouraged to act as unadvertised sales reps for Ring.