On Wednesday, the EU Parliament approved the Artificial Intelligence Act, a rule aimed at guaranteeing that fast-changing technology remains safe and in compliance with fundamental human rights while simultaneously promoting innovation.
According to a news statement on the parliament’s website, MEPs approved the law with 523 votes in favour, 46 against, and 49 abstentions after reaching an agreement with member states in December.
“Europe is NOW a global standard-setter in AI,” tweeted Thierry Breton, the EU Commissioner for the Internal Market, on X (previously Twitter).
According to the paper, the legislation categorises technology based on risk, ranging from “unacceptable” — which would prohibit some applications — to high, medium, and low hazard.
The new regulations prohibit certain AI applications that endanger people’ rights, such as biometric classification systems based on sensitive features and untargeted scraping of facial photos from the internet or CCTV footage to generate facial recognition databases. Emotion recognition in the workplace and in schools, social scoring, predictive policing (when based only on profiling or measuring a person’s features), and artificial intelligence that manipulates human behaviour or exploits people’s vulnerabilities will all be prohibited.
The AI Act also prevents law enforcement from using biometric identification systems (RBI) except in strictly defined conditions. ‘Real-time’ RBI can only be used if certain protections are followed.
“We finally have the world’s first binding law on artificial intelligence, to reduce risks, create opportunities, combat discrimination, and bring transparency,” Internal Market Committee co-rapporteur Brando Benifei stated during the plenary debate on Tuesday. He emphasised that undesirable AI activities will be forbidden in the EU, while workers’ and citizens’ rights will be protected.
The legislation is anticipated to go into effect at the end of the legislative session in May, after clearing final checks and gaining approval from the European Council.
The EU AI Act comes as global concerns grow about the potential for technology abuse, including the threat of ‘deepfakes’ or other types of artificial intelligence that manufacture phoney events, such as images and movies. Some governments, like China and India, have issued guidelines for governing artificial intelligence. Some US towns and states have also implemented legislation regulating the use of technology in specific areas, such as police investigations and hiring.
