Hungarian Conservative

EU Parliament Adopts Legislation on AI Usage to Protect Rights and Ensure Accountability

Pixabay
The new rules prohibit AI applications that could violate citizens’ rights. These include biometric categorization based on sensitive personal data or the creation of facial recognition databases using images downloaded from the internet or closed-circuit television networks.

On Wednesday, the European Parliament adopted legislation on artificial intelligence, which it says is aimed at protecting fundamental rights, democracy, the rule of law, and environmental sustainability against high-risk artificial intelligence (AI) and promoting innovation.

The new rules prohibit AI applications that could violate citizens’ rights. These include biometric categorization based on sensitive personal data or the creation of facial recognition databases using images downloaded from the internet or closed-circuit television networks. They also prohibit the use of emotion recognition systems in workplaces and schools, the AI-based risk categorization of citizens based on their social affiliation or background, and the use of systems capable of influencing people’s behaviour or exploiting their vulnerabilities.

Remote biometric identification (RBI) systems, with a few exceptions, may not be used by law enforcement agencies. Real-time RBI can now only be used under strict safeguards, such as temporal and geographical limitations, with separate judicial or regulatory authorization. It may be used, for example, in searching for missing persons or preventing terrorist attacks. The retrospective use of such systems, following the determination of a criminal offence, is considered high risk and subject to judicial authorization.

Clear obligations also apply to other high-risk artificial intelligence systems. These include critical infrastructures, education and training, employment, basic private and public services, law enforcement, migration, border management, justice, and certain democratic processes. When using such systems, risks must be assessed and mitigated, logs must be kept of their use, and human oversight must be ensured. Citizens will have the right to complain about decisions affecting their rights based on high-risk artificial intelligence systems and receive explanations. General-purpose artificial intelligence systems and the AI models they use must meet transparency requirements,

including detailed summaries of EU copyright and education-related content.

Additional requirements apply to higher-performance AI models posing system risks, including model evaluations, assessment and mitigation of system risks, and reporting of incidents. According to the new legislation, regulatory test environments must be established in EU member states, and tests must be conducted under real-world conditions, involving small and medium-sized enterprises and startups to ensure that mature and trained innovative artificial intelligence systems enter the market.

The final text of the legislation still needs to be officially approved by the Council of the EU, after which it will be published in the Official Journal of the EU. The regulation will enter into force on the twentieth day following publication.

‘This is an important step today, as we conclude a legislative process that establishes the legal framework for the use of artificial intelligence, serving the interests of EU citizens and businesses,’

Fidesz MEP Edina Tóth commented on the adoption of the legislation. According to Balázs Hidvéghi, also and MEP for Fidesz, the new rules help address citizens’ fears of artificial intelligence and increase people’s trust in AI.

The EP statement emphasized that a human-centred regulation has been developed, focusing on consumer protection while allowing for innovation and the exploitation of economic opportunities. ‘It has always been important for us that EU regulation on artificial intelligence, while placing the safety of individuals at the centre and addressing the risks associated with the use of AI, also provides fair competition conditions for digital SMEs,’ the MEPs said.

Edina Tóth added: ‘It is important that the development and application of AI within Europe remain under human supervision, be safe, transparent, non-discriminatory, and serve the well-being of society and the environment. I trust that the adopted regulation will encourage investments and innovation in the sector in Hungary.’


Related articles:

Hungary at the Forefront of Embracing the EU’s Artificial Intelligence Regulation
European Parliament Votes on Regulating Artificial Intelligence

Sources: Hungarian Conservative/EP

The new rules prohibit AI applications that could violate citizens’ rights. These include biometric categorization based on sensitive personal data or the creation of facial recognition databases using images downloaded from the internet or closed-circuit television networks.

CITATION