It’s checkout time at the supermarket, you’re standing there with your favourite tipple – but the member of staff who needs to confirm your age has mysteriously disappeared. With the UK population consuming around 1.7 billion bottles of wine in 2020 alone, this is a problem many of us have faced, especially now that there are 51% more self-service checkouts in supermarkets than there were five years ago.

But what if a potential drinker could simply look into a camera and have his or her age automatically verified by an algorithm? That’s exactly what age-recognition technology from Yoti, a London-based software company that has already performed millions of checks for age-restricted online services, promises.

However, as Yoti’s CEO outlined in a recent letter to the Minister for Science, Innovation and Technology, stress-free checkouts cannot happen overnight. As the recipients of Robin Tombs’ letter suggest, one problem is legislation. Current UK legislation was passed before machine learning was a real thing, and restricts the use of age-recognition technology.

Combined with stakeholder concerns, an automated future suddenly seems less immediate – although the technology’s enormous social and technological benefits certainly mean it will eventually arrive on the high street.

Age Detection 101

Like other examples of advanced AI, Yoti’s age detection technology is based on data. Millions of facial photos are taken, anonymized except for the month and year of birth. The pattern of the face is then converted into numbers.

Based on this, an algorithm can use pattern analysis to calculate the approximate age of anyone who looks into a Yoti camera – even newbies who fancy a cigarette or beer.

This is not to say that the system can determine birthdays precisely – and it doesn’t have to. “If I’m a liquor salesman,” explains Sean O’Neill of Gartner in the US context, “I need to make sure you’re over 21. I don’t really need to know your exact age.”

Content from our partners

Under these conditions, Tomb’s “facial age estimation” platform appears to be working well. For example, the technology was deployed in several UK supermarkets as part of a Home Office trial in 2022, accompanied by a second Yoti product focused on digital IDs for edge cases. In the 98,000 age checks carried out as part of the trial, not a single underage drinker circumvented the rules. So the trial was not perfect. A government post-mortem said that while the rollout of age estimation technology suggests there is “appetite” for automating ID checks, the trial showed that the system was still sensitive to environmental factors such as bright light “that could affect reliability”.

If, as Tombs argues, the test provided “really positive” evidence of the robustness of age verification technology, such systems certainly have other practical benefits. Aside from making the shopping process smoother – especially when age verification accounts for up to 50% of self-checkout interactions – the absence of physical staff arguably makes shopping safer.

As O’Neill points out, a full 5% of physical assaults faced by shop staff are due to disputes related to age verification, although the number of attacks on staff in England and Wales increased by half in 2022-23.

Legislative challenges

Considering that age verification technology allows workers to focus on more productive tasks, it’s no wonder the sector is thriving. This is true for Yoti too. The Insight Partners found that the global identity verification market could reach nearly $32 billion by 2030.

But why, despite all this frenzy of activity, is this technology not now widely used in UK shops? Read Tombs’ letter to MP Peter Kyle and you will quickly get the answer. One challenge, it quickly becomes clear, is legal.

The Licensing Act 2003 requires a “responsible person” to authorise the sale of alcohol, which is an obvious problem with computerised solutions such as Yoti’s. Other regulations from 2014 require shoppers to carry proof of age with a holographic or ultraviolet security feature.

“Right now,” Tombs says, “that virtually rules out using a digital ID on a smartphone or using your face” – although in practice, programs like Challenge 25 mean that employees habitually make educated guesses about a person’s age.

For Tombs, the reason for this discrepancy is simple: technology has overtaken technocrats. “We had our first discussion with the Home Office about reusable digital IDs in 2016,” he says. “At that time, we were not discussing facial age verification – because we had not yet developed that solution.”

Now that age detection technology has matured, Parliament can hopefully catch up, and certainly references in the King’s Speech to a new Digital and Smart Data Bill could herald the improvements Yoti is seeking.

Not that Tombs and his team are necessarily destined to go their own way. In addition to Labour’s ambivalence towards ambitious new technologies, public perception is another obstacle. “Everyone is afraid of Big Brother,” O’Neill puts it eloquently, and Yoti bemoans the occasional confusion between their anonymised platform and end-to-end, totalitarian-style facial recognition.

Yoti is doing its best to allay concerns. For one, any images its cameras capture are quickly deleted. For another, the images remain locally on individual devices rather than being dumped online. But when even Ofcom’s CEO is publicly expressing scepticism about the technology and barely half of people are aware of facial recognition in general, evangelists still have a long way to go.

Nevertheless, Tombs remains optimistic about the future. Not least given the incentives that companies have to remain honest – if you steal a customer’s image, he warns, you ruin your reputation – he believes that change could come by the end of this legislative session.

The work of major corporations like Walmart and jurisdictions like Jersey certainly suggests he may be right. That’s certainly good news for workers and drinkers alike.