Sainsbury’s shopper felt like a ‘criminal’ after facial recognition error
A data strategist has revealed she felt like a “criminal” after Sainsbury’s staff used facial recognition software to mistakenly identify her as a criminal, leading to her being removed from the store.
Warren Rajah, 42, of Elephant and Castle, south London, was shopping at his local branch on January 27 when he was approached by staff, asked to leave and his purchases confiscated.
‘Distraught’ Mr Rajah questioned the decision as staff pointed to a sign indicating the store was using facial recognition technology.
It was later revealed that he was confused with another person who was listed as a criminal in the system and was in the store at the time.
Sainsbury’s later apologized to Rajah and stated that there were no problems with the Facewatch technology currently used in seven of its stores.
On being misidentified, Mr Rajah told the Press Association: “You feel terrible, you feel like a criminal and you don’t even understand why.”
He added: “Telling you to leave the store without any explanation gives you the impression that you have done something wrong.
“If you talk to anyone in public, that’s what they’ll tell you; when you’re pushed and excluded from an environment, you automatically think you’ve done something wrong, especially when it comes to security.
“It’s just a normal human reaction.”
Mr Rajah said he contacted Facewatch after he was escorted out of the store, and Facewatch told him he was not in its database after sending him a copy of his passport and his own photograph.
Sainsbury’s later apologized and offered her a £75 shopping voucher.
A spokesperson for the firm said: “We have contacted Mr Rajah to sincerely apologize for his experience at our Elephant and Castle store.
“This was not an issue with the facial recognition technology used, but rather a case of approaching the wrong person in the store.”
Britain’s second-largest supermarket chain said the technology was part of efforts to detect shoplifters and curb a sharp rise in retail crime in recent years.
The website states that the system has a “99.98 percent accuracy rate and each alert is reviewed by trained peers before any action is taken.”
Sainsbury’s said the system issued an alert based on criminal behavior sent by the store or other nearby retailers using Facewatch.
Mr Rajah said he had “no interest” in shopping at Sainsbury’s anymore and wanted people to be aware of the facial recognition technology used in stores.
He said: “This also borders on fascist, how can something be done to you and not have understanding? How can you be excluded from a space and not have understanding, explanation?”
A Facewatch spokesperson said: “We are sorry to hear about Mr Rajah’s experience and understand why this may be upsetting. “This incident resulted from an incident of human error when a member of staff in store approached the wrong customer.
“Our data protection team followed the usual process to verify his identity and confirmed that he was not on our database and was not subject to any alerts generated by Facewatch.”
They added that if someone requests access to the subject, the data is not stored or used for any other purpose and is deleted after the person proves who they are.
Jasleen Chaggar of Big Brother Watch said: “The idea that we are all just one facial recognition error away from being falsely accused of a crime or kicked out of a store without explanation is utterly chilling.
“To add insult to injury, innocent people seeking relief must jump through hoops and hand over more personal data to discover what they are accused of.
“When companies are finally forced to admit that technology has done wrong, in many cases they are offered nothing but an apology.”
Ms Chaggar said the organization “regularly hears from members of the public who are traumatized after being accidentally caught in this privatized biometric surveillance network”.
The Information Commissioner’s Office (ICO) said: “Facial recognition technology can help retailers detect and prevent crime and provides clear public interest benefits. However, its use must comply with data protection law.”
“Retailers should carefully consider the risks of misidentification and have robust procedures in place to ensure the accuracy and integrity of the personal information they collect and process.
“This is particularly important where personal information is used in situations that could have a serious impact on an individual.”
