In the ultra-high-tech digitised world we live in, it’s almost anomalous that there are still things we can’t buy without permission from a living, breathing, flesh-and-bones human being. How quaint and old-fashioned!
Yet that’s the reality of purchasing alcohol, tobacco, vapes, fireworks and plenty more age-restricted products. And it poses a problem for self-service. A bottle of wine in your supermarket basket means the self-checkout experience is not entirely autonomous. You know you will have to get approval from a member of staff in the middle of scanning all your other items. It’s a point of friction many shoppers would prefer to avoid.
Retailers agree. In the UK, off-trade alcohol sales alone are worth £27bn a year, tobacco around £6.5bn, vapes £2.5bn, and lottery ticket sales in excess of £8bn. Age-restricted products are big business. But selling them carries risks. A mistake by a member of staff that leads to someone under age buying a restricted item can land you in hot water with the authorities. Carrying out checks takes time and may lead to delays and conflict.
For all of these reasons, the emergence of automated age verification that can take the place of human approval at self-service kiosks is very welcome for both shoppers and retailers. But how does it work, and how reliable is the technology?
How automated age checks work
The latest generation of on-kiosk age verification systems rely on AI. The customer looks into a camera and the AI software examines facial characteristics to calculate whether they are comfortably above the required age threshold. This is usually set well above the legal purchasing threshold, as per the Challenge 25 scheme for alcohol.
If the AI decides the customer looks old enough, the system allows the sale to proceed. If there is any doubt, the check can be escalated to require verification by a member of staff. This could also be automated using document scanners. If the customer cannot present an acceptable form of identification that proves they are the right age, the transaction is refused.
Fans of AI age estimation point out that it is not the same as identifying someone, so is more acceptable from a privacy perspective. But that feels like a moot point given that a) age verification for age-restricted products has always been carried out by asking for ID, and b) in cases where the system doesn’t verify a customer as being the right age, they are subsequently asked to present ID anyway.
It’s likely that this close link between age verification and identification will be strengthened further as biometric technology continues to evolve. If we reach the point where digital payments are authorised via a fingerprint or facial scan rather than by a card, for example, then it would be logical to wrap age verification in with that as a single process. The buyer’s personal details, including their date of birth, would be held on their billing account records anyway. And confirming age by identifying the buyer is more failsafe than even the best AI estimations.
Faster for customers, safer for staff
The immediate benefit of automated age verification is increased speed and efficiency. Linking back to the supermarket self-checkout experience, instead of waiting for a busy member of staff to come round and dial in the approval code for that bottle wine, an AI estimation can approve the sale almost instantly.
That also frees up employees. Rather than moving constantly between checkout alerts, staff deployed to monitor self-checkout can focus on issues and customer queries that require more attention.
Automated checks can also reduce mistakes caused by fatigue, distraction or subjective judgement. AI tools are not foolproof. But they are relentlessly consistent in applying rules and parameters. They can also support stronger compliance controls by creating an auditable record of all approvals and refusals.
The overall outcome is a win-win for retailers and their customers. Automated age verification takes a repetitive and sometimes contentious task out of the hands of staff, while simultaneously helping to lower the risk of illegal sales, fines and reputational damage. For customers, it means a faster, smoother and more consistent experience.
