The United Nations (UN) has explained that certain artificial intelligence systems are in its sights. In particular, it considers that facial recognition systems can represent a very serious risk in the non-respect of human rights, especially when this technology is used in public spaces. This is why the UN calls for a moratorium on certain uses of AI.

Facial recognition in the UN’s sights

The respect of human rights around the world is one of the first concerns of the UN, which considers that facial recognition can hinder these international rights. Michelle Bachelet, the UN High Commissioner for Human Rights, said:

“All countries, without exception, must expressly ban artificial intelligence applications that do not respect international human rights law. […] Their use can have negative, even catastrophic, effects if they are used without sufficient consideration of how they affect the rights of individuals.”

As part of its work on technology and human rights, the Office of the United Nations High Commissioner for Human Rights (OHCHR) today released a report analyzing how AI, including profiling, automated decision-making and other machine learning technologies, affects people’s right to privacy and other rights, including the rights to health, education, freedom of movement, freedom of peaceful assembly and association, and freedom of expression.

A call for a moratorium on certain artificial intelligence systems

The intergovernmental organization is calling for a moratorium on these technologies and more specifically on facial recognition. Michelle Bachelet adds:

“Given the continued rapid growth of AI, closing the huge accountability gap in the way data is collected, stored, shared and used is one of the most pressing human rights issues we face. The risk of discrimination from decisions made by AI – decisions that can change, define or damage human lives – is very real. This is why systematic assessment and monitoring of the effects of AI systems is needed to identify and mitigate human rights risks.”

For example, the UN considers that governmental “social rating” systems should be banned altogether. In addition, again according to the UN, the inferences, predictions and monitoring carried out by AI tools, including the search for information on human behaviour patterns, raise serious questions.

The AI report highlighted several AI malfunctions in the past that have allegedly prevented people from receiving welfare benefits or getting jobs. Some innocent people have allegedly been arrested by the police based on poorly trained facial recognition systems that have not correctly recognized people with African physical features, for example.

Without preaching to any one nation, the UN’s message is addressed to all policy makers. According to Ms. Bachelet, “we must act now to put in place human rights protections for the use of AI, for the good of all.

Translated from L’ONU appelle à un moratoire sur l’utilisation de la reconnaissance faciale