Sainsbury's pauses facial recognition at East Dulwich store after wrongful ejection

Sainsbury's has suspended its use of live facial recognition technology at its East Dulwich superstore after staff wrongly challenged a customer as a shoplifter. The incident occurred on 6 August when Matt Arnold, a 46-year-old comedy promoter, was buying supplies for an event at Dulwich Hamlet Football Club.

Arnold was stopped at a self-service till while waiting for a staff member to approve an alcohol purchase. He told BBC London that two managers approached him and said he had to leave, claiming he had been identified by the store’s AI system for an incident earlier in the week.

Read Next in News London1 min read

Sevenoaks Road Orpington closed due to emergency incident

Traffic disruption is expected following the closure of the major road in south east London.

A shoplifter does not walk around with that much shopping, they don't scan it through, they don't put their Nectar card through. But what upset me was thinking this is what the future could be, people just listen to what the machine tells them to do without thinking about the consequences.

Matt Arnold, comedy promoter

Arnold said he was escorted off the premises and later saw a CCTV monitor displaying a red circle around his face. Sainsbury's contacted him the following day to apologise, attributing the mistake to human error rather than a failure of the Facewatch system.

System accuracy disputed

Related Borough Coverage3 min read

5,500 London children join Dolly Parton's book scheme

The free book-gifting programme has expanded to seven boroughs, with more than 100,000 children now eligible.

Sainsbury's said the Facewatch system has a 99.98% accuracy rate and that every match is reviewed by a trained manager. A spokesperson for the supermarket told the BBC that the technology was not at fault in Arnold’s case, adding that initial trials across two stores had resulted in a 46% reduction in logged incidents of theft and anti-social behaviour.

Facewatch, the company supplying the technology, also said its system was not to blame. A spokesperson said a correct alert was sent to the store but was mishandled by staff. The company described the suspension as a precautionary measure while employees undergo additional training.

Arnold questioned the system’s reliability, arguing that even a 0.2% error rate could affect thousands of customers given the number of people who shop at Sainsbury's each week. He said he had donated a £150 goodwill voucher from the supermarket to a local food bank.

Calls to scrap the technology

Silkie Carlo, director of civil liberties group Big Brother Watch, called for Sainsbury's to abandon the technology. She said the incident demonstrated the risks of treating customers like criminals and argued that mistakes were inevitable when retailers conducted indiscriminate ID checks using surveillance systems.

Arnold said he was concerned about the pressure on staff to follow automated alerts without applying human judgement. He told the BBC that while his experience was distressing, he felt sympathy for employees who were not properly equipped to use the system in a busy store environment.

Sainsbury's cited rising violence against retail workers as a reason for deploying the technology. The British Retail Consortium reported 1,600 daily incidents of violence and abuse against shopworkers in 2023, up from 455 a day in 2019-20. The supermarket said over 90% of offenders flagged by the system did not return to its stores.

What happens next

Sainsbury's has paused the facial recognition system at the East Dulwich store while it investigates the incident and retrains staff. Facewatch said the suspension was a precautionary step to prevent further alerts during the review period. It is unclear when the system will be reinstated or whether Sainsbury's will extend the pause to other branches.

Arnold has called for the technology to be suspended across all Sainsbury's stores. The supermarket has not indicated whether it will consider a wider review of the Facewatch system.