International. In a resolution adopted with 377 votes in favour, 248 against and 62 abstentions, MEPs warn of the biases incurred by algorithms and emphasise the need to incorporate human supervision and legal controls to avoid discrimination, in particular when using artificial intelligence in border controls.
"Final decisions must always be taken by people," the MEPs stress, "and people subject to these types of tools must be able to appeal their decisions."
Concern about the risk of discrimination
The resolution emphasizes that many algorithm-based identification technologies make more mistakes in identifying and classifying people who are racialized or belonging to certain ethnic communities, LGBTI people, children and the elderly, and also women.
To ensure respect for fundamental rights, MEPs call for algorithms to be transparent, traceable and leave a documentary footprint. Whenever possible, authorities should use open source programs to also contribute to transparency.
Controversial technologies
Arguing the issue of privacy and human dignity, MEPs urge a permanent ban on automatic recognition systems in public spaces, and make it clear that citizens should only be subjected to surveillance if they are suspected of having committed a crime. Private facial recognition databases (such as Clearview, which already works) and predictive techniques based on behavioral data should also be banned, adds the EP (European Parliament).
MEPs also reject citizens' social scoring systems which are given a classification on the basis of their behaviour or personality.
Parliament is also concerned about the use of biometric data for the remote identification of people. The text refers in particular to the iBorderCtrl project, an "intelligent lie detection system" that has been used in tests in Hungary, Latvia and Greece in the framework of the control of entries into European territory. MEPs encourage the Commission to initiate infringement proceedings against Member States if necessary.
Statement by the rapporteur
Petar Vitanov (S&D, Bulgaria) said: "There are no conditions for respect for fundamental rights. For the first time, we are calling for a moratorium on the deployment of facial recognition systems in the police field, as the technology has proven ineffective and often discriminatory. We are against ai-based behavioral prediction techniques, as well as the processing of biometric data for mass surveillance. It's a victory for the citizens."
Source: European Parliament.
