By Bertrand Beauté

The year is 1989. In dark theatres around the world, astonished filmgoers watch a vision of the 21st century as pictured by Robert Zemeckis and Bob Gale, the directors of the Back to the Future trilogy. Alongside the flying cars, hoverboards, self-drying clothes, robot maids and other innovations that haven’t yet become reality, the second instalment in the series did predict the omnipresence of biometrics. Sent forward into 2015 by the legendary DeLorean, the film’s heroes open the door to their house, provide identification to the police and pay for a taxi all by placing their fingers on a digital fingerprint scanner.

A future dreamed up in the 1980s is now becoming a reality, as biometrics become increasingly significant in our lives. Many companies now use digital fingerprints and facial recognition to allow employees access to their buildings, and many airports have also adopted this technology. Most smartphones now unlock by pressing a finger to a sensor or by looking at the camera. And let’s not forget that biometric passports are now the norm in more than 150 countries.

"The use of biometric systems is booming. They are increasingly present in our day-to-day lives," says Christophe Remillet, CEO of OneVisage, a Swiss startup that specialises in facial recognition. Worth an estimated $41.08 billion in 2023, the global biometrics market is expected to reach $150.58 billion by 2030, or an annual growth of 20% during that time, according to predictions from consulting firm Global View Research. Biometrics Research Group agrees, believing that the market will be worth $77.9 billion in 2026. 

"We are irrefutably moving towards more and more biometrics in order to simplify identification and improve security"

Christophe Remillet, CEO of OneVisage

Why are we seeing such a boom? "The rise of biometrics systems is inherently linked to the failure of other identification systems (passwords, PINs, OTPs)," says Remillet. "If you look at passwords, for example, each person has to manage dozens. They all need to be complex and different for maximum security, but in reality, people use the same password across many accounts and generally the combinations are extremely simple and can be easily hacked. We are irrefutably moving towards more and more biometrics in order to simplify identification and improve security."

Of course, all the digital giants (Google, Amazon,  Facebook, Apple, Microsoft) have  developed biometric systems for their products. But there are also a myriad of companies that specialise in biometrics, an industry that can be split into three parts: sensors, recognition software and cloud services.

In concrete terms, biometrics – which literally means "measure of life" – refers to all processes that can identify a person using a biological or behavioural component. While facial and fingerprint recognition are the best known and most widespread biometrics, and also the most likely to be used in the future, there are many others such as recognition via voice, iris, veins, body odour, and even keyboard typing rhythm. To simplify telephone communication with clients, Migros Bank has used an automatic voice recognition system since 2020. PostFinance has done the same since 2018 and Swisscom started in 2021. Customers who choose to use the service can supply a sample of their voice which will be used in future calls to formally identify them. This method makes it easier to identify customers by removing the traditional questions such as date of birth and account number. According to Swiss newspaper Le Matin Dimanche, PostFinance has gathered approximately 1.35 million voiceprints, or more than half of all its customers.

"There are many applications for bio metrics, which opens up a myriad of opportunities," says Laetitia  Ramelet, project manager at  TA-SWISS, a foundation that conducted a large study on the risks and opportunities of biometrics in 2022. "In addition to replacing passwords to unlock our smartphones, biometrics could eventually be used in all services that require personal identification."

For example, on 20 July,  Amazon announced the rollout of its new palm recognition payment system, known as Amazon One, in all shops of US chain Whole Foods Market, which it acquired in 2017. Chinese internet giant Tencent and French electronic payment company Worldline are testing palm payments as well.

Biometrics are also being used more often in the humanitarian sector. In Ukraine, for example, facial recognition can be used to search for missing persons or to identify people who have died. And new applications of this technology will arise in the future, particularly in healthcare. "Machines are capable of determining much more than someone’s identity from their face, voice or words. They are able to make conclusions about emotions and physical and mental states," adds Ramelet. "In medicine, recognition of biometric signatures can help with the early detection of diseases such as Parkinson’s, Alzheimer’s and depression."

According to its advocates, biometrics technology has two major advantages: simplicity of use and security. "With biometrics, you don’t need a code or software hidden somewhere on a chip. You’re paying with your identity," said Gilles Grapinet, CEO of Worldline, in an interview with BFM TV. So that’s simplicity. And as for security: "All hackers that try to steal passwords will be blocked," says Christophe  Remillet. "Phishing attacks will no longer happen."

That said, biometric technology is not infallible itself. In early 2023, for example, a British journalist at the magazine Vice was able to trick his bank’s voice recognition system by using free voice AI technology. Facial and fingerprint recognition software have also been found wanting. "There is no such thing as infallible techno logy," said researcher Sébastien Marcel, head of the Biometrics Security and Privacy research group, in an interview with Swissquote.

"I’m eagerly awaiting the adoption of more restrictive legislation because many companies are using biometrics in ways that are ill-advised"

Christophe Remillet, CEO of OneVisage

Furthermore, the development of biometrics also raises concerns – many fear that widespread biometrics will be less like Back to the Future 2 and more like the dystopia of 1984, the famous novel by George Orwell in which the entire population is under permanent surveillance. To avoid that pitfall, in its report, TA-SWISS recommends that Switzerland implements a precise legal framework. Christophe Remillet agrees: "I’m not interested in a future like 1984. I’m eagerly awaiting the adoption of more restrictive legislation because many companies are using biometrics in ways that are ill-advised. As a result, people are scared of being under permanent surveillance." However, the CEO of OneVisage does not believe that these fears should halt the progress made in biometrics. "Technology is advancing. Today, no one refuses to wear a seatbelt. It will be the same with biometrics. In three to five years, it’ll be widespread and no one will question these security systems."

Authentication or identification?

Biometrics systems perform two main roles: authentication and identification. Authentication is simpler: think unlocking a smartphone or accessing buildings. To configure an authentication system, the first step is for users to save their biometric data (a fingerprint or face scan for consumer electronics). This data will be stored and used as a reference. Any time the system is accessed after the initial setup – unlocking a smartphone, for example – it compares the data collected by the sensor (fingerprint scanner or camera) against the data stored in the database. If it’s a match, the device unlocks. If not, it stays locked. For authentication, the bio metrics system needs to answer ‘yes’ or ‘no’ by comparing collected data to stored data. All authentication systems are made up of three essential components: the sensor that collects biometric data, a storage location for the original data, and software to compare the two.

Identification is much more complex: the system has to provide the name of an unknown person based on biometric data. For example, in a police investigation, detectives can capture the face of a person in a surveillance video. That face would then be compared to a database, such as a database of faces of wanted people. In this case, the system doesn’t answer ‘yes’ or ‘no’ but rather gives a match probability. This process is not without considerable risk: in 2020, Robert Williams, an African-American, spent 30 hours in custody because the facial recognition software used by Detroit police determined that the photo from his driver’s licence was the same as the image of a watch thief caught on surveillance cameras.