Mastercard’s newly launched face recognition payment system is already raising accuracy concerns

• Earlier this week, Mastercard launched a new payment system - ‘smile to pay’.
• Users can make a payment by showing their face or palm, instead of swiping a card.
• Third-party authentication and data usage raise security concerns over new payment system.

Mastercard recently launched a new facial recognition payment system that lets shoppers make payments with their face or hand gestures. This new payment system is being rolled out to biometric payment systems including fingerprint scanning and facial recognition.

The new payment system from Mastercard is known as ‘smile to pay.’ It is aimed to reduce transaction times, shorten queues in shops, increase security and improve hygiene by going cardless. Currently, this feature is live in Brazil and Sao Paulo. The company says it plans to roll it out globally later this year.

With this new tech company aims to change the way we pay but it also raises concerns relating to data storage, customer privacy and crime risks.

To use facial recognition-based payments, Mastercard’s biometric checkout system will provide biometric authentication with third-party companies. The company has already partnered with Payface, NEC, Fujitsu, Aurus, PopID and PayByFace.

According to reports, customers have to install a third-party app that will take their pictures and payment information. All the information will be stored on third-party servers. While checkout, the customer's face will be matched with the stored data to make a payment.

Where does concern lie?
When a third-party app is involved, privacy protection concern is a major issue. We can assume while giving authentication to a third party, customers will require to give consent. But the question arises, will customers know what they are agreeing to?

In the end, Mastercard and biometric service providers will decide how they are going to use data and who can access it.

Accuracy and Bias
Earlier, the accuracy of facial recognition has been challenged. According to the National Institute of Standards and Technology, the best facial authentication algorithms have a 0.08% chance of error. Till now, we don’t know how accurate is Mastercard’s ‘smile to pay’ feature.

According to a NIST study, out of 189 facial recognition, the majority of them were biased meaning they were less accurate on people among ethnic minorities and racial groups. Mastercard has not revealed to which extent it has overcome these challenges.

Also, there are no future plans disclosed by Mastercard to deal with situations if technology misidentifies a person. For example, if person A is recognised as person B, then money would be debited from the wrong person.

Mastercard study shows that 74% of people are in favour of using the facial recognition payment system. But we can see the report doesn't have any foolproof evidence to support the survey.

SEE ALSO:
This new tech can detect lies within minutes by analyzing involuntary eye movement
Hero Electric scooter catches fire in Odisha, no casualties or injuries were reported