The demand for biometrics to reduce retail crime continues to rise, but the risk of flawed deployment of the technology can also be seen in an unfortunate incident and allegations of a massive invasion of privacy.
According to Retail Week, the British Retail Consortium is urging the authorities to allocate more resources to reduce retail theft as part of its “Get tough on Retail Crime” campaign. It’s not just theft the BRC is concerned about, noting that incidents of abuse and violence against retail staff have increased from 870 a day in 2021-2022 to 1,300 across the UK in 2022-2023.
Elections for candidates for the post of police and crime commissioner will be held on 2 May.
Just last week, the government announced a £55.5 million investment in facial recognition over the next four years, specifically aimed at reducing shoplifting and abuse of shop assistants.
Corsight Artificial Intelligence released a video showing when someone who previously stole from a retailer used facial recognition to detect when they re-entered the premises.
Retailers can use Corsights’ software to add facial images captured from shoplifters to their watch lists. When the person steals again, the software alerts the staff. The faces of people not on the watch list are blurred in the video to protect their privacy.
Ofer Ronen, the company’s executive vice president of global business development, noted in an interview with American Security Magazine that the company has also developed an app for retail facial recognition and analysis. He suggested that the company’s “facial intelligence” could be used to identify when customers are waiting to buy high-value items that are locked away.
The parent company of New World Supermarkets in New Zealand has apologised after a Maori woman was mistaken for a shoplifter and confronted by staff, according to New Zealand news reports. The woman also said she did not know the store was using facial recognition technology until after she was confronted.
Foodstuff’s facial recognition pilot quickly proved controversial and is under investigation by New Zealand’s Privacy Commissioner.
The retailer uses a facial recognition system and a process of manual review by two trained people to confirm or reject a candidate’s match.
A spokesman for the food company said: “In this case, our team did not err on the side of caution and erred.”
This false match may indicate that the biometric system used is not as effective for Maori and other darker-skinned people.
Interestingly, a spokesperson for the Privacy Commissioner’s Office made an overgeneralization mistake when commenting on facial recognition and demographic differences: “Global assessments of the most accurate facial recognition software indicate that people of color, especially women of color, are more likely to have false matches.”
NIST’s global assessment of demographic differences actually found that most algorithms produced demographic differences or biases in matching results, with the most accurate algorithms showing little or no evidence of bias.
Meanwhile, in the US, Target has been accused of failing to obtain informed consent from customers before using facial recognition loss prevention systems to process biometric information.
The class-action lawsuit was filed in Illinois under the state’s Biometric Information Privacy Act (BIPA), and according to NBC Chicago, employees of the retail giant have openly discussed its facial recognition system online. Walmart operates 14 investigative centers and two forensic LABS in the United States that process video and “analyze fingerprints,” according to the complaint.
Leave feedback about this