Shoppers have reported being wrongly identified as thieves and removed from stores after being flagged by facial recognition systems, according to an investigation by The Guardian.

The newspaper said it had spoken to multiple people who claimed they were falsely accused after being linked to the Facewatch system, a live facial recognition platform used by a growing number of retailers to help tackle shoplifting and retail crime.

Facewatch is currently used by retailers including B&M, Home Bargains, Sports Direct, Farmfoods and Spar. The company says its system has a 99.98 per cent accuracy rate and sent more than 50,000 alerts of “known offenders” to shops last month.

However, the investigation raised concerns over the process for shoppers who believe they have been wrongly identified, with some saying they struggled to understand why they had been approached, how their data was being used, or how they could challenge the decision.

The cases included shoppers being asked to leave stores or being monitored by security staff after being linked to Facewatch alerts. In some cases, people later discovered they had either been wrongly associated with alleged past incidents or had been misidentified by store staff.

Privacy campaigners and biometrics experts have warned that oversight of the technology has not kept pace with its rollout across both the retail sector and policing.

The Home Office has previously acknowledged that facial recognition cameras are more likely to incorrectly identify black and Asian people than white people, and women more than men.

A Sainsbury’s spokesperson told The Guardian: “We have sincerely apologised to [an alleged shoplifter] for his experience in our Elephant and Castle store. This was not an issue with the facial recognition technology in use but a case of the wrong person being approached in store.

“The Facewatch system has a 99.98 per cent accuracy rate and all matches are reviewed by trained managers, with additional training provided after this incident to ensure our safeguards are consistently followed.”

Facewatch chief executive Nick Fisher said the cases referenced by The Guardian related to “human error in the way processes were carried out in-store”, rather than a failure of the technology itself.

He said: “We are sorry these individuals experienced being challenged while shopping and understand why this would have been upsetting.

“These three errors are extremely rare cases when viewed in the context of the more than 500,000 alerts we send to retailers each year, but we recognise that any mistake is upsetting for the individual concerned. The system is designed to support, not replace, human decision-making.”

The Information Commissioner’s Office said facial recognition technology must be used carefully and transparently.

A spokesperson said: “We recognise the harm and upset that can be caused by misidentification. For this reason, use of facial recognition technology must strictly comply with data protection law and be handled with care and transparency.

“If someone has concerns about how their data has been collected, used, or shared, and those concerns cannot be resolved with the retailer directly, they have the right to raise a complaint with us.

“We also continue to actively regulate in this area and will be publishing further retail-focused guidance to support retailers in understanding and meeting their data protection obligations, while ensuring the public is properly protected.”

Click here to sign up to Retail Gazette‘s free daily email newsletter