Pretty faces

While debates on setting ethical limits to AI are going on, more established technologies expand without much ethical considerations. "Technology without morality is barbarous; morality without technology is impotent," said Feeman Dyson, a 20th century British-American theoretical physicist and mathematician. Like many technologies facial recognition can be the best or the worst thing depending on how it is used. Facial recognition technologies are supported by industry leaders such as Idemia or Thales, and also smaller, more dedicated, companies such as Veriff or iProov.

Facial recognition is already accepted in a one-to-one context by pretty much everyone: many of us unlock their phone by looking at it several times a day without a second thought. Then, facial recognition based on a voluntary use and linked with an ID is a widely accepted authentication method: eGates in airports around the world control that your face matches the one stored in your ePassport chip. Facial recognition is often used in authentication in various contexts, for instance, starting in July 2023, financial institutions in Thailand will be using facial recognition for customers conducting transactions over THB 50,000 (EUR 1,330) or more via mobile banking.

Facial recognition and AI are combined to provide an easy way to implement non-intrusive age control. Latest analysis tools by Neurotechnology, for instance, include age estimation. Facial age estimation, developed by UK-based startup Yoti, is also used by Meta when users are joining Facebook Dating, a Tinder-like function of Facebook, available only to adult users.

Biometric payments based on face recognition are also increasingly commonplace, especially in Asia. While it started in 2017 in China, the adoption of facial recognition based payments in physical stores received a boost with the coronavirus pandemic; payment systems are widely available and fueled by the local payment leaders: Alipay and WeChat. In Japan, Yahoo! Mart is introducing a self-service point-of-sale (POS) cash register that allows customers to pay via facial recognition. Of course, these systems require prior enrolment when ID verification is completed. Also Mastercard launched their “Biometric Checkout” system, which uses biometrics, face recognition or fingerprint, as the core means of identification for payment. As they require prior enrolment and acceptance by the user, facial recognition uses in mass transit systems especially in some cities in China and in Moscow, Russia, are in the same category.

When facial recognition becomes a one-to-many database extraction, its uses become way more questionable. Unsurprisingly, China and Russia are at the forefront of using massive facial recognition as government surveillance means. For instance, cameras, which are everywhere in Xinjiang, are used by the government to have an individual control of each person in the population, according to a report by the UK Foreign, Commonwealth & Development Office. If these policies were benchmarked in Xinjiang, they are now expanding to the whole country, including Tibet, but also large cities like Beijing or Shanghai, as demonstrated by Wall Street Journal reporters Josh Chin and Liza Lin in their book titled "Surveillance state." Similarly, the Russian government is using facial recognition as a means of population surveillance and control. Moscow is home to one of the world’s largest video surveillance networks, consisting of over 160,000 cameras, including more than 3,000 connected to a facial recognition system. since the invasion of Ukraine by Russia in February 2022, facial recognition technology is being used as a preventive measure to prevent protests, leading to the arrest of the Kremlin’s opponents, according to Biometric Update and Reuters.

The western world is going along different paths on this point. In the US, the New York Supreme Court ruled in favor of Amnesty International and the Surveillance Technology Oversight Project (STOP) and ordered the New York City Police Department (NYPD) to disclose thousands of records of how it obtained and used facial recognition technologies against Black Lives Matter protesters between March and September 2020.

The Italian Data Protection Agency just issued a statement banning the use of facial-recognition technology and similar biometrics systems until either a specific law is adopted or at least until the end of next year.

At the same time, in the context of the upcoming 2024 Summer Olympics, France is passing a law formalizing the introduction of algorithmic video surveillance. The technology constantly identifies, analyzes, classifies the bodies, physical attributes, gestures, body shapes, gait, which are unquestionably biometric data, says privacy watchdog La Quadrature du Net. "France is once again violating EU’s law, consecrating its title of Europe surveillance’s champion," they add. The French Senate adopted a draft law on testing facial recognition technology in public spaces. The law will allow judicial investigators and intelligence services to use remote biometric identification in public for three years. "The new draft specifies that real-time facial recognition use in public will be limited to tracking down terrorists by intelligence services, child abductions and particularly serious crimes," according to Biometric Update. Earlier this year, the French Constitutional Council, allowed algorithmic processing of video feeds during the 2024 Paris Summer Olympics and Paralympics.

So, where do we stand in the technology vs. morality debate?