In a conversation with my teenage daughter last week, I pointed out news report which expressed concerns about the use of face recognition technology in several school canteens in North Ayrshire, Scotland. Nine schools in the area have recently launched this practice as a means of paying for lunch faster and minimizing the risk of COVID, although since it’s on pause technology implementation.
When I asked my daughter if she had any concerns about using face recognition technology in her school canteen, she casually replied, “Not really. However, this would make things much faster when paying. “
Her words confirm the concern of the children much less aware of their data rights compared to adults. And although there are special regulations and precautions for children under a number data protection legislation, the use of facial recognition technology can pose unique privacy risks.
Face recognition technologies identifies and verifies the identity of people by detecting, photographing and comparing persons with images from a database. The technologies are powered by artificial intelligence (AI), in particular the technology known as machine learning.
Machine learning predicts results based on historical data or algorithms that have been submitted to the system. So for facial recognition, machine learning predicts the identity associated with the digital representation of a person’s face, or “face imprint”, based on a database of facial images. The software adapts through this experience, and over time it learns to generate predictions more easily.
Face recognition technology is now used in a variety of ways, such as to authenticate employees, to unlock personal smartphones, to tag people on social media platforms like Facebook, and even to monitoring objectives in some countries.
Face recognition technology is not a problem in itself. Rather, the question is how it is used and, in this case, the fact that technology has already penetrated school corridors and targeted a vulnerable demographic group: children.
So what are the privacy issues for children?
Your facial imprint is your data, so for any facial recognition system, it’s important to understand how image databases are collected and stored. Although I may be reluctant to agree to the use of face recognition technology to enter a concert venue, I would not be excited if my fingerprint was retained for “other commercial purposes of the company” (a phrase that is appears quite often in the fine print of ticket sales in connection with the use of personal data).
If face recognition technology is used in a school setting, we will need clear information on whether and how students’ images will be used outside the purpose of the lunch queue. For example, will they be shared with third parties and for what purpose? Problems can arise, for example, if the person’s fingerprints are related to other data about the child, such as his preferences for lunch. Third parties could theoretically use this data for marketing purposes.
We will also need information on how the images will be protected. If students’ fingerprints are not properly protected or the system is not stable enough to repel hackers, this creates cybersecurity risks. It may be possible for hackers to link child fingerprints to other data about them and track them.
The increased risk of confidentiality associated with the use of face recognition technologies in schools is also linked to informed consent. Although the UK data protection law specifies that children aged 13 and over may agree to the processing of their personal data, this does not mean that they fully understand the consequences. For example, one survey found that children between the ages of eight and 15 had difficulty understanding Instagram’s terms and conditions.
Children, parents and guardians should be provided with nothing less than complete information in a language that children can easily understand. Every data subject, including a child, has the right to know exactly how his or her personal data will be processed, shared and stored and can indicate the conditions under which their consent will apply. Anything less than caution and transparency will risk jeopardizing children’s privacy.
Normalize children’s observation?
These are just some of the issues raised by the use of face recognition technologies in schools. Face recognition technology poses other risks, such as errors that can lead to incorrect student billing. And as with any AI system, we need to worry about whether the algorithms and datasets are free from biasand have clean, complete and representative training data.
It is important that the use of face recognition technologies in schools also contributes to the normalization of children’s observation. It is possible that the knowledge they are tracked in this way will affect the well-being of some children.
Not surprisingly, the UK Data Observatory, the Office of the Commissioner for Information, has intervene to explore the use of face recognition technology in school lunch queues. And in the light of the investigation, it is gratifying to see that the North Ayrshire Council has it. stopped deployment practice.
But as the digital age progresses, it is possible that the use of face recognition technology among students will resume and even become more widespread. If this happens, the use of facial recognition should bring significantly more benefits than risks, taking into account the special circumstances of using the technology on children.