Latest news with #BrandoBenifei
Euronews
26-03-2025
- Politics
- Euronews
Exclusive: Hungary's Gay Pride surveillance would breach the EU's AI Act, says leading MEP
ADVERTISEMENT Digital facial recognition systems Hungary proposes to use to trace participants of Budapest's Gay Pride march in Budapest will violate the EU's AI Act and data protection rules, according to an Italian MEP charged with monitoring the AI law. Last Tuesday Hungary's Parliament ruled to ban the march - which usually takes place in June - in a modification to rules affecting public assembly from being held outside, and threatened those flouting the ban with fines of 200,000 Forint (€500), reserving the right use digital facial recognition tools to identify those in breach. According to MEP Brando Benifei (Italy/Socialists & Democrats) - a co-chair of the European Parliament's AI monitoring group, tasked with overseeing implementation of the AI Act - this would be a clear breach of the EU's new AI Act. "What was announced by Orbán is illegal: it is forbidden by the AI Act to use biometric cameras to identify protestors. You cannot surveil people peacefully protesting unless they are linked to serious crimes, like homicide, terrorism," Benifei told Euronews. First big test for landmark law, says Benifei Benifei, who also steered the AI Act through the Parliament as a rapporteur, added that businesses may only sell facial recognition systems in Europe subject to these provisions, and otherwise risk losing their license to do so. The MEP urged the European Commission to look into the issue. "We are still collecting data and are still in the early phase of implementing the AI Act. But Hungary will be the first big test for us and the Commission should be very clear on this," he added. Tamás Lattmann, an EU law expert in Hungary, agreed that the Hungarian regulation violates the AI Act in a post on social media specifying those crimes which justify the use of AI tools. "In every other case their use is forbidden," Lattmann said. Benifei added that - regardless of the implementation of the AI Act - he considered use of the facial recognition technology in the context of the march would also be a clear breach of the EU's General Data Protection Regulation. Distinction drawn between real-time, and post event facial analysis A Commission spokesperson told Euronews that the assessment of legality would depend under the AI Act would depend on whether Hungarian Police use the facial recognition software in real-time or after the event, for so-called post-processing. Use of real-time biometric identification of protesters is prohibited by the law unless there is an imminent terrorist threat, search for missing persons or investigation of serious crimes with prior judicial approval, the spokesperson said. "Remote biometric identification, including facial recognition in public spaces, is a high-risk application, that requires transparency, accountability, and fundamental rights impact assessments," the spokesperson said, referring to rules affecting post-event identification coming into effect in August, after the scheduled march. Euronews has asked the Hungarian government for comment on its proposed use of the technology. In Hungary, the National Expertise and Research Center has the right to analyze facial data. This organization has collected biometric data of citizens since 2015 from ID cards and driving licenses. On request by the police, an algorithm searches the database for matches. Hungarian media reports have suggested that the changes to the public assembly rules might open the door to facial recognition being used to control issues such as traffic offenses and littering. However, experts have also questioned whether the administration has the technical resources to even surveil the pride march. ADVERTISEMENT The EU's AI Act – which regulates AI tools according to the risk they pose to society – came into force last August. Its provisions started to apply gradually: the act will be fully in force in 2027. Each EU member state will have to set up a national regulator to monitor companies' compliance of the AI Act by August of this year. They will collaborate with the European Commission's AI Office, which supports the national authorities.
Euronews
14-02-2025
- Business
- Euronews
EU Parliament plans AI and workplace initiative for next year
The European Parliament will begin negotiations on an own-initiative legislative proposal to tackle issues around artificial intelligence and the workplace in the first quarter of this year, with the aim to adopt a report in early 2026, sources from within the institution told Euronews. Last year, the bloc's AI Act – stringent rules which regulate AI systems according to the risk they pose to society – entered into force, and the Parliament has been vocal about the need for additional rules to cover workplace changes arising as a result of AI technology. Lawmaker Brando Benifei (Italy/S&D), the MEP in charge of determining the Parliament's position on the AI Act, repeatedly sai d that possible AI and workplace rules could cover issues such as obligations to involve social partners in bargaining, as well as safety and training elements. The AI Act already bans specific practices to protect workers, such as social scoring or real-time remote biometric identification. In addition, the EU institutions reached an agreement on platform work last year, with the aim to protect and improve the working conditions of people working in the gig economy. Possible legislation The issue was flagged in the mission letter of Roxana Mînzatu, the EU Commissioner in charge of Social Rights and Skills, Quality Jobs and Preparedness. 'I would also like you to focus on the impact of digitalisation in the world of work. This should notably be done through an initiative on algorithmic management and through possible legislation on AI in the workplace, following consultation with social partners. You will also propose to introduce a right to disconnect,' the mission letter, drafted by Commission President Ursula von der Leyen, said. The European Commission did not comment on when such a proposal could be presented. A survey published on Thursday by the EU executive found that most Europeans support the use of AI in the workplace. More than 60% of Europeans view robots and AI in the workplace positively and more than 70% believe that they improve productivity, the survey suggested. 'While a majority supports the use of robots and AI to make decisions at work, 84% of Europeans believe that AI needs to be carefully managed to protect privacy and ensure transparency in the workplace,' it said. 'Artificial intelligence is now an integral part of modern workplaces. We need to use it in a way that supports and protects workers. We have a solid foundation of rules, with our General Data Protection Regulation, the AI Act and the Platform Work Directive. We will review how these rules are applied to address the impact of AI across the labour market,' Mînzatu said.
